30 November 2018 in IEEE Transactions on Radiation and Plasma Medical Sciences
17 May 2018 in IEEE Transactions on Radiation and Plasma Medical Sciences
Monolithic crystals are considered as an alternative for complex segmented scintillator arrays in Positron Emission Tomography (PET) systems. Monoliths provide high sensitivity, good timing and energy resolution while being cheaper than highly segmented arrays. Furthermore, monoliths enable intrinsic depth of interaction capabilities and good spatial resolutions mostly based on statistical calibrations. To widely translate monoliths into clinical applications, a time-efficient calibration method and a positioning algorithm implementable in system architecture such as FPGAs are required. We present a novel positioning algorithm based on Gradient Tree Boosting (GTB) and a fast fan beam calibration requiring less than 1h per detector block. GTB is a supervised machine learning technique building a set of sequential binary decisions (decision trees). The algorithm handles different sets of input features, their combinations and partially missing data. GTB models are strongly adaptable influencing both the positioning performance and the memory requirement of trained positioning models. For an FPGA-implementation, the memory requirement is the limiting aspect. We demonstrate a general optimization and propose two different optimization scenarios: One without compromising on positioning performance and one optimizing the positioning performance for a given memory restriction. For a 12mm high LYSO-block, we achieve a spatial resolution better than 1.4mm FWHM.
8 April 2024 in Nature Biomedical Engineering
The clinical prospects of cancer nanomedicines depend on effective patient stratification. Here we report the identification of predictive biomarkers of the accumulation of nanomedicines in tumour tissue. By using supervised machine learning on data of the accumulation of nanomedicines in tumour models in mice, we identified the densities of blood vessels and of tumour-associated macrophages as key predictive features. On the basis of these two features, we derived a biomarker score correlating with the concentration of liposomal doxorubicin in tumours and validated it in three syngeneic tumour models in immunocompetent mice and in four cell-line-derived and six patient-derived tumour xenografts in mice. The score effectively discriminated tumours according to the accumulation of nanomedicines (high versus low), with an area under the receiver operating characteristic curve of 0.91. Histopathological assessment of 30 tumour specimens from patients and of 28 corresponding primary tumour biopsies confirmed the score’s effectiveness in predicting the tumour accumulation of liposomal doxorubicin. Biomarkers of the tumour accumulation of nanomedicines may aid the stratification of patients in clinical trials of cancer nanomedicines.
20 October 2023 in IEEE Transactions on Neural Networks and Learning Systems
Artificial intelligence (AI) is entering medical imaging, mainly enhancing image reconstruction. Nevertheless, improvements throughout the entire processing, from signal detection to computation, potentially offer significant benefits. This work presents a novel and versatile approach to detector optimization using machine learning (ML) and residual physics. We apply the concept to positron emission tomography (PET), intending to improve the coincidence time resolution (CTR). PET visualizes metabolic processes in the body by detecting photons with scintillation detectors. Improved CTR performance offers the advantage of reducing radioactive dose exposure for patients. Modern PET detectors with sophisticated concepts and read-out topologies represent complex physical and electronic systems requiring dedicated calibration techniques. Traditional methods primarily depend on analytical formulations successfully describing the main detector characteristics. However, when accounting for higher-order effects, additional complexities arise matching theoretical models to experimental reality. Our work addresses this challenge by combining traditional calibration with AI and residual physics, presenting a highly promising approach. We present a residual physics-based strategy using gradient tree boosting and physics-guided data generation. The explainable AI framework SHapley Additive exPlanations (SHAPs) was used to identify known physical effects with learned patterns. In addition, the models were tested against basic physical laws. We were able to improve the CTR significantly (more than 20%) for clinically relevant detectors of 19height, reaching CTRs of 185(450–550 keV).
9 January 2023 in Physics in Medicine & Biology
Objective. Positron emission tomography (PET) detectors providing attractive coincidence time resolutions (CTRs) offer time-of-flight information, resulting in an improved signal-to-noise ratio of the PET image. In applications with photosensor arrays that employ timestampers for individual channels, timestamps typically are not time synchronized, introducing time skews due to different signal pathways. The scintillator topology and transportation of the scintillation light might provoke further skews. If not accounted for these effects, the achievable CTR deteriorates. We studied a convex timing calibration based on a matrix equation. In this work, we extended the calibration concept to arbitrary structures targeting different aspects of the time skews and focusing on optimizing the CTR performance for detector characterization. The radiation source distribution, the stability of the estimations, and the energy dependence of calibration data are subject to the analysis. Approach. A coincidence setup, equipped with a semi-monolithic detector comprising 8 LYSO slabs, each 3.9 mm × 31.9 mm × 19.0 mm, and a one-to-one coupled detector with 8 × 8 LYSO segments of 3.9 mm × 3.9 mm × 19.0 mm volume is used. Both scintillators utilize a dSiPM (DPC3200-22-44, Philips Digital Photon Counting) operated in first photon trigger. The calibration was also conducted with solely one-to-one coupled detectors and extrapolated for a slab-only setup. Main results. All analyzed hyperparameters show a strong influence on the calibration. Using multiple radiation positions improved the skew estimation. The statistical significance of the calibration dataset and the utilized energy window was of great importance. Compared to a one-to-one coupled detector pair achieving CTRs of 224 ps the slab detector configuration reached CTRs down to 222 ps, demonstrating that slabs can compete with a clinically used segmented detector design. Significance. This is the first work that systematically studies the influence of hyperparameters on skew estimation and proposes an extension to arbitrary calibration structures (e.g. scintillator volumes) of a known calibration technique.
23 January 2023 in IEEE Transactions on Radiation and Plasma Medical Sciences
In modern large-scale PET systems, transferring the digitized raw detector data at high count rates to a centralized processing unit is a challenge. Processing data on FPGAs close to the detectors can reduce data early on and improve scalability of the PET system. We present and evaluate an FPGA implementation of gradient tree boosting (GTB) for one-dimensional position estimation of gamma interactions in the scintillator. GTB is a supervised machine learning algorithm based on building ensembles of binary decision trees. Models were trained offline and inferred in an FPGA (XC7K410T-2FFG676 Kintex-7). Input features and GTB parameters influencing both positioning performance and model size were varied while evaluating the inferred models concerning data throughput and FPGA resource consumption as well as positioning performance. We achieved throughputs per detector between 2.94 × 106 and 4.55 × 106 gamma interactions per second. For an optimized GTB model, resource consumption could be reduced by factors of 17 and 10 to less than 1% (2.51 × 103 look-up tables) of available logic and 1.26% (20 BRAMs) of memory resources, while maintaining a positioning performance of 98.63% when compared to the model with the best positioning performance. The presented framework can be easily adapted to other photosensors and scintillator configurations.