ABSTRACT
PURPOSE: To develop and validate an Artificial Intelligence (AI) model based on texture analysis of high-resolution T2 weighted MR images able 1) to predict pathologic Complete Response (CR) and 2) to identify non-responders (NR) among patients with locally-advanced rectal cancer (LARC) after receiving neoadjuvant chemoradiotherapy (CRT). METHOD: Fifty-five consecutive patients with LARC were retrospectively enrolled in this study. Patients underwent 3 T Magnetic Resonance Imaging (MRI) acquiring T2-weighted images before, during and after CRT. All patients underwent complete surgical resection and histopathology was the gold standard. Textural features were automatically extracted using an open-source software. A sub-set of statistically significant textural features was selected and two AI models were built by training a Random Forest (RF) classifier on 28 patients (training cohort). Model performances were estimated on 27 patients (validation cohort) using a ROC curve and a decision curve analysis. RESULTS: Sixteen of 55 patients achieved CR. The AI model for CR classification showed good discrimination power with mean area under the receiver operating curve (AUC) of 0.86 (95% CI: 0.70, 0.94) in the validation cohort. The discriminatory power for the NR classification showed a mean AUC of 0.83 (95% CI: 0.71,0.92). Decision curve analysis confirmed higher net patient benefit when using AI models compared to standard-of-care. CONCLUSIONS: AI models based on textural features of MR images of patients with LARC may help to identify patients who will show CR at the end of treatment and those who will not respond to therapy (NR) at an early stage of the treatment.
Subject(s)
Artificial Intelligence , Chemoradiotherapy, Adjuvant/methods , Magnetic Resonance Imaging/methods , Neoadjuvant Therapy/methods , Rectal Neoplasms/diagnostic imaging , Rectal Neoplasms/therapy , Aged , Aged, 80 and over , Cohort Studies , Female , Humans , Male , Middle Aged , Prospective Studies , ROC Curve , Rectal Neoplasms/pathology , Rectum/diagnostic imaging , Rectum/pathology , Reproducibility of Results , Retrospective Studies , Treatment OutcomeABSTRACT
Proton and carbon ion beams are used in the clinical practice for external radiotherapy treatments achieving, for selected indications, promising and superior clinical results with respect to x-ray based radiotherapy. Other ions, like [Formula: see text] have recently been considered as projectiles in particle therapy centres and might represent a good compromise between the linear energy transfer and the radiobiological effectiveness of [Formula: see text] ion and proton beams, allowing improved tumour control probability and minimising normal tissue complication probability. All the currently used p, [Formula: see text] and [Formula: see text] ion beams allow achieving sharp dose gradients on the boundary of the target volume, however the accurate dose delivery is sensitive to the patient positioning and to anatomical variations with respect to photon therapy. This requires beam range and/or dose release measurement during patient irradiation and therefore the development of dedicated monitoring techniques. All the proposed methods make use of the secondary radiation created by the beam interaction with the patient and, in particular, in the case of [Formula: see text] ion beams are also able to exploit the significant charged radiation component. Measurements performed to characterise the charged secondary radiation created by [Formula: see text] and [Formula: see text] particle therapy beams are reported. Charged secondary yields, energy spectra and emission profiles produced in a poly-methyl methacrylate (PMMA) target by [Formula: see text] and [Formula: see text] beams of different therapeutic energies were measured at 60° and 90° with respect to the primary beam direction. The secondary yield of protons produced along the primary beam path in a PMMA target was obtained. The energy spectra of charged secondaries were obtained from time-of-flight information, whereas the emission profiles were reconstructed exploiting tracking detector information. The obtained measurements are in agreement with results reported in the literature and suggests the feasibility of range monitoring based on charged secondary particle detection: the implications for particle therapy monitoring applications are also discussed.
Subject(s)
Heavy Ion Radiotherapy/adverse effects , Helium/adverse effects , Polymethyl Methacrylate/radiation effects , Radiation Monitoring/methods , Radiotherapy Planning, Computer-Assisted/methods , Dose-Response Relationship, Radiation , Humans , Scattering, RadiationABSTRACT
Charged particle beams are used in particle therapy (PT) to treat oncological patients due to their selective dose deposition in tissues with respect to the photons and electrons used in conventional radiotherapy. Heavy (Z > 1) PT beams can additionally be exploited for their high biological effectiveness in killing cancer cells. Nowadays, protons and carbon ions are used in PT clinical routines. Recently, interest in the potential application of helium and oxygen beams has been growing. With respect to protons, such beams are characterized by their reduced multiple scattering inside the body, increased linear energy transfer, relative biological effectiveness and oxygen enhancement ratio. The precision of PT demands online dose monitoring techniques, crucial to improving the quality assurance of any treatment: possible patient mis-positioning and biological tissue changes with respect to the planning CT scan could negatively affect the outcome of the therapy. The beam range confined in the irradiated target can be monitored thanks to the neutral or charged secondary radiation emitted by the interactions of hadron beams with matter. Among these secondary products, prompt photons are produced by nuclear de-excitation processes, and at present, different dose monitoring and beam range verification techniques based on prompt-γ detection are being proposed. It is hence of importance to perform γ yield measurement in therapeutic-like conditions. In this paper we report on the yields of prompt photons produced by the interaction of helium, carbon and oxygen ion beams with a poly-methyl methacrylate (PMMA) beam stopping target. The measurements were performed at the Heidelberg Ion-Beam Therapy Center (HIT) with beams of different energies. An LYSO scintillator, placed at [Formula: see text] and [Formula: see text] with respect to the beam direction, was used as the photon detector. The obtained γ yields for the carbon ion beams are compared with results from the literature, while no other results from helium and oxygen beams have been published yet. A discussion on the expected resolution of a slit camera detector is presented, demonstrating the feasibility of a prompt-γ-based monitoring technique for PT treatments using helium, carbon and oxygen ion beams.
Subject(s)
Heavy Ion Radiotherapy/methods , Photons , Polymethyl Methacrylate/radiation effects , Scintillation Counting/methods , Carbon/chemistry , Carbon/therapeutic use , Heavy Ion Radiotherapy/adverse effects , Heavy Ion Radiotherapy/standards , Helium/chemistry , Helium/therapeutic use , Humans , Linear Energy Transfer , Proton Therapy , Relative Biological Effectiveness , Scintillation Counting/instrumentationABSTRACT
Nowadays there is a growing interest in particle therapy treatments exploiting light ion beams against tumors due to their enhanced relative biological effectiveness and high space selectivity. In particular promising results are obtained by the use of 4He projectiles. Unlike the treatments performed using protons, the beam ions can undergo a fragmentation process when interacting with the atomic nuclei in the patient body. In this paper the results of measurements performed at the Heidelberg Ion-Beam Therapy center are reported. For the first time the absolute fluxes and the energy spectra of the fragments-protons, deuterons, and tritons-produced by 4He ion beams of 102, 125 and 145 MeV u-1 energies on a poly-methyl methacrylate target were evaluated at different angles. The obtained results are particularly relevant in view of the necessary optimization and review of the treatment planning software being developed for clinical use of 4He beams in clinical routine and the relative bench-marking of Monte Carlo algorithm predictions.