The use of hadrons for cancer treatment is increasing worldwide due to their more favorable physical and biological properties over photon or electron beams. As of February 2025, more than 120 proton and 17 carbon-ion facilities are treating patients all over the world, and more than 30 are being constructed (Particle Therapy Co-Operative Group, 2025). However, the radiobiological effectiveness of ion beams can change significantly as they degrade in matter; in addition, carbon ions can undergo nuclear reactions which give rise to a build-up of secondary fragments travelling beyond the range of the primary beam. The clinical field at treatment depth comprises multiple components with varying biological effectiveness. This brings about the need for fast and accurate in-phantom monitoring not only of the physical dose, but also of additional physical quantities which are related to radiobiological effectiveness (the so-called radiation quality, determined by the type and energy spectrum of particles at the treatment depth). This is true not only for treatments with charged hadrons such as protons or carbon ions, but also for emerging binary therapies such as Boron Neutron Capture Therapy (BNCT).

Microdosimetry is a spectrometric technique which measures energy imparted by ionizing radiation in sensitive volumes of micrometric dimensions. It can provide an effective approach for radiation quality monitoring in hadron therapy: starting from microdosimetric quantities it is possible to derive radiation quality descriptors of clinical interest, such as the Linear Energy Transfer (LET) (Kellerer, 1972; Braby et al., 2023) or a physics-based estimation of the Relative Biological Effectiveness (RBE). The latter is based either on specific weighting functions, derived from iterative unfolding of radiobiological and microdosimetric measurements carried out in the same radiation field (Loncol et al., 1994; Tilikidis et al., 1996; Parisi et al., 2020), or on radiobiological models such as the Microdosimetric Kinetic Model and its various modifications (Hawkins, 2003; Bellinzona et al., 2021).

Microdosimetric measurements provide a fast characterization of the radiation field, which could be performed as a routine procedure. For this reason, its introduction into Quality Assurance (QA) procedures for hadron therapy is gaining increasing interest and has been recently recommended by the International Commission of Radiation Units (Braby et al., 2023). However, this requires the availability of validated instruments for the measurement of microdosimetric quantities.

Reference detectors for microdosimetry are the Tissue-Equivalent Proportional Counters (TEPCs), gas-filled counters working in the proportionality regime where the chemical composition of both the detector walls and the filling gas is chosen to be as similar as possible to that of human tissue. This ensures that both the mass stopping power and the secondary particle spectrum in the detector materials are approximately the same as those in tissue. The gas pressure inside the cavity is set by applying a density scaling principle: its thickness in mass per area must be the same as that of a micrometric tissue volume with density 1 g/cm³.

TEPCs generally have a spherical or cylindrical shape, with a cathode shell made of conductive plastic (typically A150, which is muscle-equivalent) and a central metallic anode wire. The filling gas is usually a methane- or propane-based tissue-equivalent mixture. If the voltage difference between cathode and anode is such that the counter works in proportional mode, each particle crossing the sensitive volume generates a signal pulse with an amplitude proportional to the initial ionization yield. The latter is then converted to energy imparted by means of a calibration factor, assuming that the W-value (mean energy to create an ion pair) does not depend on particle type and energy.

Commercial versions of TEPCs for microdosimetry were available on the market, most notably the FWT-LET1/2 counter manufactured by Far West Technology, Inc. This detector has a spherical sensitive volume with a diameter of 1.27 cm. The FWT-LET1/2 has been used in some clinical centers for a microdosimetric characterization of the radiation field, both for ion therapy (Coutrakon et al., 1997; Kase et al., 2006; Martino et al., 2010) and BNCT applications (Endo et al., 2004; Hu et al., 2020).

The main challenge in the development of microdosimetric counters for particle therapy is the high intensity of the radiation field. Both in therapy with charged hadrons and in BNCT, the incident flux is of the order of 10⁷–10¹⁰ cm⁻² s⁻¹. The cross-sectional area of detectors developed for these applications must therefore be small enough to cope with such a high fluence rate without significant pile-up effects. To achieve this goal, two strategies have been developed: either to miniaturize TEPCs as much as possible, down to cavity sizes less than 1 mm in diameter, or to move to solid-state technology, with silicon- or diamond-based devices of physical micrometric size. Given its large cross area, measurements with the FWT-LET1/2 counter can be done using a particle fluence rate orders of magnitude lower than the therapeutic one.

The Legnaro National Laboratories of INFN have extensive experience in developing miniaturized microdosimeters for hadron therapy. Since 1990, several devices were developed for research purposes, with varying cavity diameter, electrode design and external size. The first prototypes had a cylindrical cavity with a diameter and height of 1 mm and an A150 cathode with a thickness of 6 mm, for an overall external size of 15 mm (Cesari et al., 2001). One prototype was built with field tubes and another without, to quantify the impact of field distortions at the margins of the sensitive volume. The response of the two detectors was found to be almost the same (De Nardo et al., 2004a).

A second generation of prototypes was built with the aim of minimizing the counter external size. The smallest ones have a sensitive volume diameter of only 0.9 mm and an external diameter of 2.7 mm. The A150 cathode wall was 0.35 mm thick and surrounded by a Rexolite insulator. These detectors have been used to perform microdosimetric measurements at several clinical facilities, both in proton (De Nardo et al., 2004b; De Nardo et al., 2004c) and carbon-ion beams (Colautti et al., 2018a; Colautti et al., 2018b). The most recent version was designed with enlarged gas ducts to be operated without gas flow: when filled with pure propane, it was shown to provide a stable response for at least 1 year (Conte et al., 2019). The use of this gas instead of a tissue-equivalent mixture does not alter significantly the shape of microdosimetric spectra, if a proper scaling factor is applied (Chiriotti et al., 2015).

The new engineered mini-TEPCs developed at LNL have several advantages over previous prototypes, making them attractive for use in clinical settings. The sensitive volume size of 1 mm is comparable to the dimensions of voxels used in treatment planning and allows measurements at fluence rates up to 10⁷ cm⁻² s⁻¹ without severe pile-up distortions. The custom low-noise preamplifier has a dynamic range of more than four orders of magnitude, allowing to measure events in the entire lineal energy range of interest for both hadron therapy and BNCT. Since the anode and guard tubes are set to zero bias through the preamplifier, a high-voltage power supply with only one negative channel is needed, with very low power requirements (less than 100 mW).

The production time and overall cost have been significantly reduced thanks to serialized construction of components, improving also reproducibility of response between different sensors. The simplified assembly also allows for the replacement of individual detector components in case of failure (Bianchi et al., 2024). The detector case including front-end electronics can fit in a box with dimensions of 25 × 21 × 5 cm³ (see Figure 1), for an overall encumbrance comparable to silicon-based microdosimeters. The total weight of less than 1 kg and the sealed detector design allow easy transportation to clinical centers.

From a clinical perspective, the tissue-equivalent composition of TEPCs is an advantage over solid-state devices especially in mixed radiation fields: non-equivalent materials would require specific conversion factors for each particle type and energy, even though a unique average factor is generally applied with silicon or diamond microdosimeters. Mixed fields are always present in particle therapy, particularly beyond the Bragg peak or in out-of-field regions, where fragments generated by nuclear reactions can change significantly the radiation quality. This is also the case in BNCT, where the tissue-equivalence of the neutron converter is of critical importance for a correct assessment of all dose components.

Given their favorable features (low detection threshold, high dynamic range, tissue-equivalent response and high portability), LNL mini-TEPCs could be promising detectors for introducing microdosimetry in routine QA procedures in hadron therapy, following the recent recommendation of the International Commission of Radiation Units (Braby et al., 2023). Their main drawback is their relatively large cross-sectional area, which limits their performance at fluence rates higher than 10⁷ cm⁻² s⁻¹. They can measure at clinical intensities in both carbon-ion and BNCT fields, but this is still not possible in proton beams where single-event measurements can be performed only at reduced intensity. However, other approaches can be implemented in clinical conditions, such as measuring the mean values of the distributions in a multi-event approach, similar to standard dosimetry, and applying the variance-covariance technique (Grindborg et al., 1995).