Design and evaluation of a deep learning-based automatic segmentation of maxillary and mandibular substructures using a 3D U-Net.

Background: Current segmentation approaches for radiation treatment planning in head and neck cancer patients (HNCP) typically consider the entire mandible as an organ at risk, whereas segmentation of the maxilla remains uncommon. Accurate risk assessment for osteoradionecrosis (ORN) or implant-based dental rehabilitation after radiation therapy may require a nuanced analysis of dose distribution in specific mandibular and maxillary segments. Manual segmentation is time-consuming and inconsistent, and there is no definition of jaw subsections. Materials and methods: The mandible and maxilla were divided into 12 substructures. The model was developed from 82 computed tomography (CT) scans of HNCP and adopts an encoder-decoder three-dimensional (3D) U-Net structure. The efficiency and accuracy of the automated method were compared against manual segmentation on an additional set of 20 independent CT scans. The evaluation metrics used were the Dice similarity coefficient (DSC), 95% Hausdorff distance (HD95), and surface DSC (sDSC). Results: Automated segmentations were performed in a median of 86 s, compared to manual segmentations, which took a median of 53.5 min. The median DSC per substructure ranged from 0.81 to 0.91, and the median HD95 ranged from 1.61 to 4.22. The number of artifacts did not affect these scores. The maxillary substructures showed lower metrics than the mandibular substructures. Conclusions: The jaw substructure segmentation demonstrated high accuracy, time efficiency, and promising results in CT scans with and without metal artifacts. This novel model could provide further investigation into dose relationships with ORN or dental implant failure in normal tissue complication prediction models.

Adaptive radiotherapy for breast cancer.

Research in the field of local and locoregional breast cancer radiotherapy aims to maintain excellent oncological outcomes while reducing treatment-related toxicity. Adaptive radiotherapy (ART) considers variations in target and organs at risk (OARs) anatomy occurring during the treatment course and integrates these in re-optimized treatment plans. Exploiting ART routinely in clinic may result in smaller target volumes and better OAR sparing, which may lead to reduction of acute as well as late toxicities. In this review MR-guided and CT-guided ART for breast cancer patients according to different clinical scenarios (neoadjuvant and adjuvant partial breast irradiation, whole breast, chest wall and regional nodal irradiation) are reviewed and their advantages as well as challenging aspects discussed.

Estimation of secondary cancer projected risk after partial breast irradiation at the 1.5 T MR-linac.

PURPOSE: For patients treated with partial breast irradiation (PBI), potential long-term treatment-related toxicities are important. The 1.5 T magnetic resonance guided linear accelerator (MRL) offers excellent tumor bed visualization and a daily treatment plan adaption possibility, but MRL-specific electron stream and return effects may cause increased dose deposition at air-tissue interfaces. In this study, we aimed to investigate the projected risk of radiation-induced secondary malignancies (RISM) in patients treated with PBI at the 1.5 T MRL. METHODS: Projected excess absolute risk values (EARs) for the contralateral breast, lungs, thyroid and esophagus were estimated for 11 patients treated with PBI at the MRL and compared to 11 patients treated with PBI and 11 patients treated with whole breast irradiation (WBI) at the conventional linac (CTL). All patients received 40.05 Gy in 15 fractions. For patients treated at the CTL, additional dose due to daily cone beam computed tomography (CBCT) was simulated. The t­test with Bonferroni correction was used for comparison. RESULTS: The highest projected risk for a radiation-induced secondary cancer was found for the ipsilateral lung, without significant differences between the groups. A lower contralateral breast EAR was found for MRL-PBI (EAR = 0.89) compared to CTL-PBI (EAR = 1.41, p = 0.01), whereas a lower thyroid EAR for CTL-PBI (EAR = 0.17) compared to MRL-PBI (EAR = 0.33, p = 0.03) and CTL-WBI (EAR = 0.46, p = 0.002) was observed. Nevertheless, when adding the CBCT dose no difference between thyroid EAR for CTL-PBI compared to MRL-PBI was detected. CONCLUSION: Better breast tissue visualization and the possibility for daily plan adaption make PBI at the 1.5 T MRL particularly attractive. Our simulations suggest that this treatment can be performed without additional projected risk of RISM.

Analysis of the electron-stream effect in patients treated with partial breast irradiation using the 1.5 T MR-linear accelerator.

INTRODUCTION: The hybrid magnetic resonance linear accelerator (MRL) has the potential to test novel concepts in breast cancer patients such as daily MR-guided real-time plan adaptation. Before starting clinical trials, preparatory studies for example of the MR-dependent electron stream effect (ESE) are necessary. MATERIAL AND METHODS: To prospectively investigate the ESE, data from 11 patients treated with partial breast irradiation (PBI) at the 1.5 T MRL were evaluated. A bolus was placed on the chin and in vivo dosimetry results were compared with the dose simulated by the treatment planning system (TPS). The same measurements were carried out for three patients treated at a conventional linac. Toxicity and cosmesis were evaluated. RESULTS: Median doses measured and simulated on top/ underneath the bolus were 1.91 / 0.62 Gy and 2.82 / 0.63 Gy, respectively. Median differences between calculations and measurements were 0.8 Gy and 0.1 Gy. At the conventional linac, median measured doses on top/ underneath the bolus were 0.98 and 1.37 Gy. No acute toxicity exceeding grade 2 was recorded. Cosmesis was good or excellent and patient reported outcome measures were mostly scored as none or mild. CONCLUSION: The dose due to the ESE is low, correctly predicted by the TPS and effectively minimized by a bolus.

Marker-less online MR-guided stereotactic body radiotherapy of liver metastases at a 1.5 T MR-Linac - Feasibility, workflow data and patient acceptance.

INTRODUCTION: Stereotactic body radiotherapy (SBRT) is an established ablative treatment for liver tumors with excellent local control rates. Magnetic resonance imaging guided radiotherapy (MRgRT) provides superior soft tissue contrast and may therefore facilitate a marker-less liver SBRT workflow. The goal of the present study was to investigate feasibility, workflow parameters, toxicity and patient acceptance of MRgSBRT on a 1.5 T MR-Linac. METHODS: Ten consecutive patients with liver metastases treated on a 1.5 T MR-Linac were included in this prospective trial. Tumor delineation was performed on four-dimensional computed tomography scans and both exhale triggered and free-breathing T2 MRI scans from the MR-Linac. An internal target volume based approach was applied. Organ at risk constraints were based on the UKSABR guidelines (Version 6.1). Patient acceptance regarding device specific aspects was assessed and toxicity was scored according to the common toxicity criteria of adverse events, version 5. RESULTS: Nine of ten tumors were clearly visible on the 1.5 T MR-Linac. No patient had fiducial markers placed for treatment. All patients were treated with three or five fractions. Median dose to 98% of the gross tumor volume was 38.5 Gy. The median time from "patient identity check" until "beam-off" was 31 min. Median beam on time was 9.6 min. Online MRgRT was well accepted in general and no treatment had to be interrupted on patient request. No event of symptomatic radiation induced liver disease was observed after a median follow-up of ten month (range 3-17 months). CONCLUSION: Our early experience suggests that online 1.5 T MRgSBRT of liver metastases represents a promising new non-invasive marker-free treatment modality based on high image quality, clinically reasonable in-room times and high patient acceptance. Further studies are necessary to assess clinical outcome, to validate advanced motion management and to explore the benefit of online response adaptive liver SBRT.

Development and validation of a 1.5 T MR-Linac full accelerator head and cryostat model for Monte Carlo dose simulations.

PURPOSE: To develop, implement, and validate a full 1.5 T/7 MV magnetic resonance (MR)-Linac accelerator head and cryostat model in EGSnrc for high precision dose calculations accounting for magnetic field effects that are independent from the vendor treatment planning system. METHODS: Primary electron beam parameters for the implemented model were adapted to be in accordance with measured dose profiles of the Elekta Unity (Elekta AB, Stockholm, Sweden). Parameters to be investigated were the mean electron energy as well as the Gaussian radial intensity and energy distributions. Energy tuning was done comparing depth dose profiles simulated with monoenergetic beams of varying energies to measurements. The optimum radial intensity distribution was found by varying the radial full width at half maximum (FWHM) and comparing simulated and measured lateral profiles. The influence of the energy distribution was investigated by comparing simulated lateral and depth dose profiles with varying energy spreads to measured data. Comparison of simulations and measurements was performed by calculating average and maximum local dose deviations. The model was validated recalculating a clinical intensity-modulated radiation therapy plan for the MR-Linac and comparing the resulting dose distribution with simulations from the commercial treatment planning system Monaco using the gamma criterion. RESULTS: Comparison of simulated and measured data showed that the optimum initial electron beam for MR-Linac simulations was monoenergetic with an electron energy of (7.4 ± 0.2) MeV. The optimum Gaussian radial intensity distribution has a FWHM of (2.2 ± 0.3) mm. The average relative deviations were smaller than 1% for all simulated profiles with optimum electron parameters, whereas the largest maximum deviation of 2.07% was found for the 22 × 22 cm 2 cross-plane profile. Profiles were insensitive to energy spread variations. The IMRT plan recalculated with the final MR-Linac model with optimized initial electron beam parameters showed a gamma pass rate of 99.83 % using a gamma criterion of 3%/3 mm. CONCLUSIONS: The EGSnrc MR-Linac model developed in this study showed good accordance with measurements and was successfully used to recalculate a first full clinical IMRT treatment plan. Thus, it shows the general possibility for future secondary dose calculations of full IMRT plans with EGSnrc, which needs further detailed investigations before clinical use.

Stimulation of human gamma interferon production by diterpene esters.

The diterpene ester 12-O-tetradecanoylphorbol-13-acetate (TPA) and several structurally related compounds were tested for their ability to stimulate interferon (IFN) production in primary cultures of human leukocytes. In cultures of Ficoll-Hypaque-purified mononuclear cells, TPA treatment alone induced only low levels of IFN, but TPA pretreatment of cells caused significant enhancement of IFN yields produced with phytohemagglutinin or several other T cell mitogens. In cultures of unprocessed cells derived from plateletpheresis residues or buffy coats, TPA treatment alone induced high levels of IFN and costimulation with TPA and phytohemagglutinin produced some further enhancement of IFN production. Phorbol 12,13-dibutyrate was comparable to TPA in its ability to enhance phytohemagglutinin-induced IFN production. Several other phorbol ester analogs were also active, but maximal stimulation occurred only at higher drug concentrations. Mezerein, a structurally related diterpene ester, was at least as active as TPA in stimulating IFN production in either Ficoll-Hypaque-purified or unprocessed cells. IFN produced after stimulation with TPA or mezerein, singly or in combination with phytohemagglutinin, had several properties characteristic of IFN-gamma, e.g., it was largely inactivated by dialysis at pH 2, or after exposure to sodium dodecyl sulfate, whereas it was not neutralized by antibody to IFN-alpha and IFN-beta. The stimulatory effect of diterpene esters has proved helpful in producing IFN-gamma for physicochemical analysis and other studies.

Lectins in the United States diet: a survey of lectins in commonly consumed foods and a review of the literature.

Plant lectins or phytohemagglutinins possess potent in vivo biological activities. Some, primarily of the family Leguminosae, have been shown to have deleterious nutritional effects. Little information exists, however, regarding the prevalence of lectins or the specific foods that contain lectins in the United States diet. In the present study the edible parts of 29 of 88 foods tested, including common salad ingredients, fresh fruits, roasted nuts, and processed cereals were found to possess significant lectin-like activity as assessed by hemagglutination and bacterial agglutination assays. Based on this survey and a review of the literature we conclude that dietary exposure to plant lectins is widespread. The spectrum of nutritional consequences of such exposure remains to be determined.

Chloride self exchange in Ehrlich ascites cells. Inhibition by furosemide and 4-acetamido-4'-isothiocyanostilbene-2,2'-disulfonic acid.

The effects of furosemide and 4-acetamido-4'-isothiocyanostilbene-2,2'-disulfonic acid (SITS) on steady-state Cl- flux were studied in Ehrlich mouse ascites cells. At 10 mM, furosemide inhibited isotopically-determined Cl- flux by 86% without changing cell Cl- content, indicating that influx and efflux were depressed by the same amount. These results suggest that at least 86% of the steady-state Cl- flux may occur as a one for one exchange. Half of the inhibitory effect was not reversed by vigorous washing with albumin-Ringer. A smaller portion of steady-state Cl- flux was inhibited by SITS. The maximum effect of SITS was reached near 0.6 mM; at this concentration Cl- flux was reduced by 37% without an alteration in cell Cl- content. Possible competition of environment Cl- and SITS was investigated by replacing environment Cl- with acetate or NO3. These anions reduced the efficacy of SITS because they depressed cell Cl- turnover themselves, apparently acting on the same exchange process.

Altered erythrocyte osmotic fragility in mice with Ehrlich ascites tumor.

The osmotic fragility of erythrocytes from mice with Ehrlich ascites tumor was compared with that of erythrocytes from normal control animals. Erythrocytes collected from mice 15 days after the ip injection of tumor cells exhibited a uniform pattern of abnormal resistance to hemolysis in hypotonic saline, with 50% hemolysis occurring at an average saline concentration of 0.44% compared to an average of 0.49% for 11 controls a significant difference (P less than 0.0001). Erythrocytes from mice with this tumor apparently undergo an alteration in some component of the cell membrane that regulates either permeability to cations and water or distensibility of the cell.

Identification and quantitation of solubilized I blood group substance by wheat germ agglutinin using quantitative immunoelectrophoresis.

Wheat germ agglutinin (WGA) has been shown to react specifically with solubilized I blood group substance, purified from papain treated human erythrocyte membranes. WGA and I react to form an affinity precipitate in immunodiffusion gels, a reaction which can be blocked by the incorporation of N-acetyl glucosamine into the gel. The I material was a strong inhibitor of both anti-I cold hemagglutination and WGA hemagglutination reactions. Utilizing the techniques of crossed immunoelectrophoresis we have clearly established that WGA and anti-I IgM cold antibody are reacting with the same membrane macromolecule (I antigen). WGA was then used in a rocket affinoelectrophoretic assay system to quantitate I substance. The limits of detection in this system was 25 ng.

Nature of lectin-induced alteration of potassium transfer in Ehrlich ascites tumor cells.

The way in which the lectins concanavalin A (Con A) and Ricinus communis agglutinin (Ricin) alter the K+ content of Ehrlich ascites tumor cells was investigated. Unidrectional and net fluxes were determined in unwashed cells during a time course following lectin addition. Total influx, ouabain sensitive influx, Mg++- and Na+-K+-ATPase activity were all unaffected. Cell ATP content was normal for at least 19 minutes after exposure to Con A. Early after contact with Ricin or Con A efflux was stimulated 2-3-fold, resulting in net K+ loss, but after 20 minutes efflux had returned to normal. Ricin and Con A acted similarly although Ricin was present at only 1/50 the concentration of Con A. When the findings are evaluated together with previous work it is suggested that a particular membrane glycoprotein may be concerned in the efflux alteration observed.

Interactions of lectins with plasma membrane glycoproteins of the Ehrlich ascites carcinoma cell.

Several aspects of the interaction of various lectins with the surface of Ehrlich ascites carcinoma cells are described. The order of agglutinating activity for various lectins is Ricinus communis greater than wheat germ greater than or equal to concanavalin A greater than or equal to soybean greater than Limulus polyphemus. No agglutination was noted for Ulex europaeus. Using 125I-labeled lectins it was determined that there are 1.6 and 7 times as many Ricinus communis lectin binding sites for concanavalin A and soybean lectins. Sodium deoxycholate-solubilized plasma membrane material was subjected to lectin affinity chromatography and sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The lectin receptors of the plasma membrane appeared to be heterogeneous and some qualitative differences could be discerned among the electrophoretically analyzed material, which bound to and was specifically eluted from the various lectin affinity columns. The characteristics of elution of bound material from individual lectin columns indicated secondary hydrophobic interactions between concanavalin A or wheat germ agglutinin and their respective lectin receptor molecules.

Multiple specificities of mammalian blood group substances comparatively studied with human isoagglutinins and fractionated anti-H lectins.

Purified blood group-active substances derived from different pig, horse, baboon, Rhesus monkey and human tissues were quantitatively studied for their haemagglutination inhibiting potency with: (1) human IgM anti-A and anti-B; (2) human anti-Lea and anti-Leb; (3) Ulex europaeus extracts separated into lectin fractions with respective L-fucose-inhibitable ('anti-HF') and chitobiose-cellobiose-inhibitable ('anti-HC') combining sites. Irrespective of species origin, A and B blood group activity per milligram of purified material tended to be strikingly higher in substances low in, or devoid of, Lewis blood group activity. Most of the blood group substances displayed variable but about equally balanced amounts of Ulex anti-HF and anti-HC inhibiting activity. In contrast, pig submaxillary gland mucins displayed strikingly high levels of Ulex anti-HC inihibiting activity, even in the complete absence of Ulex anti-HF inhibiting activity. These serological findings are consistent with current biochemical concepts regarding the heterosaccharide microheterogeneity of blood group-active glycoproteins.