Purpose: Daily image-guided radiotherapy (IGRT) for prostate cancer reduces treatment-related toxicity. However, positional and anatomical variations can result in patient re-setup, increased treatment time, and additional imaging dose. A simple technique of two distinct imaging modalities was investigated: initially, an anteroposterior kilovoltage (AP-kV) planar image was acquired, followed by cone beam computed tomography (CBCT). To determine whether this dual imaging modality (DIM) can reduce repeat CBCTs with an AP-kV screening image compared to CBCT alone, i.e. a single imaging modality (SIM). Methods: Patients were enrolled sequentially before and after the new departmental protocol for IGRT. Initially, all patients had a CBCT prior to treatment (SIM group) as usual. For the DIM group, AP-kV imaging was added to the first three fractions, and subsequent AP-kV imaging (on demand) for patients unable to maintain adequate bladder and bowel compliance. Sixty intact prostate or post-prostatectomy patients were included: 30 before the change in protocol (SIM group) and 30 after (DIM group). Bladder volume, rectal filling, and positioning errors were evaluated. Results: In the SIM group, 30 patients underwent a total of 1116 CBCTs. In the DIM group, 30 patients had 156 AP-kV images and 1077 CBCTs. In the SIM group, 166 were repeat CBCTs with a median of 4 repeat CBCTs per patient. In DIM group, 132 were repeat CBCTs with a median of 3 repeat CBCTs per patient. We found a significant difference in incidence of repeat CBCTs due to suboptimal bladder filling (p = 0.028) and rectal gas (p = 0.045), indicating the number of repeat CBCTs was significantly lower in patients imaged with DIM. Conclusion: The DIM technique was found to be superior to the SIM technique, as it allowed the desired bladder volume, rectal condition, and patient positioning to be achieved prior to CBCT, reducing the need for repeat CBCTs.

Combination immune checkpoint inhibitors (nivolumab and ipilimumab) are currently a first-line treatment for mesothelioma; however, not all patients respond. The efficacy of treatment is influenced by the tumor microenvironment. Murine mesothelioma tumors were irritated with various radiotherapy doses. Radiotherapy induced vasculature changes were monitored by power Doppler and photoacoustic ultrasound and analyzed via mixed-effects models. Tissue staining was used to investigate the immune cell infiltrate of tumors. The optimal radiotherapy schedule was combined with immune checkpoint inhibitors, and the survival of mice was analyzed. Using low-dose, low-fraction radiotherapy allowed favorable modification of the murine mesothelioma tumor microenvironment. Irradiating tumors with 2 Gy × 5 fractions significantly improved blood flow and reduced hypoxia, consequently increasing the presence of CD8+ and regulatory T cells in the tumor. Understanding the transient nature of these changes is crucial for optimizing the timing of therapeutic delivery. The combination of radiotherapy with dual immunotherapy (anti-PD-1 plus anti-CTLA-4) proved highly curative when administered concurrently. A diminishing rate of cures was noted with an increasing delay between radiotherapy and subsequent immunotherapy. Concurrent low-dose, low-fraction radiotherapy emerges as a translatable approach for improving the efficacy of immune checkpoint inhibitors in patients.

We present a design concept for the next generation of optical modulators. The demonstrated driver–modulator integrated silicon photonics transmitter can operate at 200 GBaud, while the power efficiency of the driver is 2.5 pJ/bit.

Artificial Intelligence (AI) based auto-segmentation has demonstrated numerous benefits to clinical radiotherapy workflows. However, the rapidly changing regulatory, research, and market environment presents challenges around selecting and evaluating the most suitable solution. To support the clinical adoption of AI auto-segmentation systems, Selection Criteria recommendations were developed to enable a holistic evaluation of vendors, considering not only raw performance but associated risks uniquely related to the clinical deployment of AI. In-house experience and key bodies of work on ethics, standards, and best practices for AI in Radiation Oncology were reviewed to inform selection criteria and evaluation strategies. A retrospective analysis using the criteria was performed across six vendors, including a quantitative assessment using five metrics (Dice, Hausdorff Distance, Average Surface Distance, Surface Dice, Added Path Length) across 20 head and neck, 20 thoracic, and 19 male pelvis patients for AI models as of March 2023. A total of 47 selection criteria were identified across seven categories. A retrospective analysis showed that overall no vendor performed exceedingly well, with systematically poor performance in Data Security & Responsibility, Vendor Support Tools, and Transparency & Ethics. In terms of raw performance, vendors varied widely from excellent to poor. As new regulations come into force and the scope of AI auto-segmentation systems adapt to clinical needs, continued interest in ensuring safe, fair, and transparent AI will persist. The selection and evaluation framework provided herein aims to promote user confidence by exploring the breadth of clinically relevant factors to support informed decision-making.

FLASH radiotherapy employs ultra-high dose rates of>40Gy s-1, which may reduce normal tissue complication as compared to conventional dose rate treatments, while still ensuring the same level of tumour control. The potential benefit this can offer to patients has been the cause of great interest within the radiation oncology community, but this has not translated to a direct understanding of the FLASH effect. The oxygen depletion and inter-track interaction hypotheses are currently the leading explanations as to the mechanisms behind FLASH, but these are still not well understood, with many questions remaining about the exact underpinnings of FLASH and the treatment parameters required to optimally induce it. Monte Carlo simulations may hold the key to unlocking the mystery behind FLASH, allowing for analysis of the underpinning mechanisms at a fundamental level, where the interactions between individual radiation particles, DNA strands and chemical species can be studied. Currently, however, there is still a great deal of disagreement in simulation findings and the importance of the different mechanisms they support. This review discusses current studies into the mechanisms of FLASH using the Monte Carlo method. The simulation parameters and results for all major investigations are provided. Discussion primarily revolves around the oxygen depletion and inter-track interactions hypotheses, though other, more novel, theories are also mentioned. A general list of recommendations for future simulations is provided, informed by the articles discussed. This review highlights some of the useful parameters and simulation methodologies that may be required to finally understand the FLASH effect.

We present a lateral tunnel epitaxy technique to grow InP membranes atop silicon waveguides on silicon-on-insulator substrates. Uniform InP membranes extending hundreds of micrometers in length can be achieved. Room temperature optically pumped lasing was also realized, showcasing the excellent crystal quality.

For the next-generation high-baud rate and low-power optical interconnects, silicon photonics has attracted widespread interest in using mature CMOS production processes to manufacture high-yield, low-cost photonic integrated circuits (PIC) and electronics integrated circuits (EIC). Using the 3D integrated silicon photonics transmitter, up to 224-Gbaud OOK signal is generated by applying a high-order partial response narrowing (HPRN) scheme and transmitted over 300-m standard single-mode fiber (SSMF) with a bit-error rate (BER) below the 20% hard-decision forward error-correction (HD-FEC) threshold of 1.5 x 10- 2 . Furthermore, for high- order modulation formats, the transmission of 260-Gb/s (130-Gbaud) PAM-4 and 300-Gb/s (120-Gbaud) PAM-6 Nyquist-shaped signals over 300-m SSMF are also realized, below the 20% HD-FEC and 25% soft-decision forward error-correction (SD-FEC) thresholds respectively. Our work implies the feasibility of low-cost optical transceivers for ultra-high-baud-rate short-reach data center optical interconnects.

We demonstrate an all-silicon universal ring resonator operating in the O-band that functions as a high-speed modulator operating at > 64 Gb/s OOK and a photodetector exhibiting up to 48 Gb/s detection rate with a responsivity of 0.3A/W.

The goal of this study was to conduct an external, independent validation of an O-(2-[18F]fluoroethyl)-l-tyrosine ([18F]FET) PET automatic segmentation network on a cohort of patients with glioblastoma. Methods: Twenty-four patients with glioblastoma were included in this study who underwent a total of 52 [18F]FET PET scans (preradiotherapy, n = 23; preradiotherapy retest, n = 9; follow-up, n = 20). Biologic tumor volume (BTV) delineation was performed by an expert nuclear medicine physician and an automatic segmentation network. Physician and automated quantitative metrics (BTV, mean tumor-to-background ratio [TBRmean], lesion SUVmean, and background SUVmean) were assessed with Pearson correlation and Bland-Altman analysis (bias, limits of agreement [LoA]). Automated and physician segmentation overlap was assessed with spatial and distance-based metrics. Results: BTV and TBRmean Pearson correlation was excellent for all time points (range, 0.92-0.98). In 2 patients with frontal lobe lesions, the network segmented the transverse sinus. Bland-Altman analysis showed network underestimation of physician-derived BTVs (absolute bias, 2.7 cm3, LoA, -13.1-18.5 cm3; relative bias, 27.9%, LoA, -95.3%-151.2%) and deviations for TBRmean were small (absolute bias, 0.03, LoA, -0.25-0.30; relative bias, 0.83%, LoA -14.27%-15.93%). Median Dice similarity coefficient, surface Dice similarity coefficient, Hausdorff distance, 95th percentile Hausdorff distance, and mean absolute surface distance were 0.83, 0.95, 10.94 mm, 3.62 mm, and 0.88 mm, respectively. Conclusion: Automated quantitative analysis was highly correlated with physician assessment; however, volume underestimation and erroneous segmentations may impact radiotherapy treatment planning and response assessment. Further training on a representative local dataset would likely be required for multicenter implementation.

We demonstrate a 3 nm t(ox) MOSCAP ring achieving a BER < 1x10(-12) up to 50 Gbps and an open eye at 100 Gbps. A 3 dB ER is achieved with 1.5 dB ILone or 1 V-pp.

We investigate the performance boundary of integrated CMOS silicon photonics transmitters by using optical equalisation techniques. Without using any off-line DSP techniques at the receiver side, experimental results demonstrate 182Gbps OOK and 308Gbps PAM-4 transmissions with the energy efficiency including driver, of better than 1pJ/bit.

Integrated electro-optic modulators offer huge potential to meet communications and computations' rapidly growing bandwidth requirements. Devices based on silicon allow high-volume, low-cost CMOS fabrication, and co-integration with the CMOS circuits. They are promising candidates for mass-producible Tb/s-scale inter-rack and intra-rack interconnects. This talk will focus on our advancement of silicon-based optical modulators: (1) miniaturized all silicon MOSCAP modulators for co-packaged optics and its integration with low voltage drivers, allowing low optical power consumption of 2 pJ/bit. (2) Novel carrier absorption enhanced electro-optical modulation in MOSCAP ring resonators towards integration with ultra-low voltage (<1V) CMOS drivers; (3) Carrier depletion ring unity device for large scale and high bandwidth density error-free links; (4) Linear DC-Kerr effect dominated silicon modulators towards lidar and quantum applications.

Optical modulators play an important role in communication systems, and silicon has been a focal point in this field thanks to its compatibility with CMOS fabrication. However, silicon's lack of inherent electro-optic behavior makes it suboptimal for modulation purposes. Conversely, potassium tantalate niobate (KTN) materials boast an improved electro-optic coefficient, presenting a path for improving modulation efficiency. However, limited research exists on KTN materials due to the difficulties associated with their fabrication. Here, a fabrication methodology is described for wafer -scale vertical integration of KTN material onto silicon -on -insulator (SOI) wafers. The resulting devices exhibit a propagation loss of 3.3 dBmm 1 and a transition loss within the range of 0.46 to 0.76 dB, which are in agreement with simulations. This method tackles the fabrication challenges and showcases the potential of utilising KTN as the integration material on silicon platform for future optical modulators.

We demonstrate a fully-integrated silicon wavelength converter utilizing four-wave mixing, featuring tunable idler filtering with >59 dB out-of-band signal suppression. Conversion of a 32-Gbaud 16-QAM signal with around 2 dB power penalty is demonstrated.

We demonstrate silicon/SiO $_{2}$ /polysilicon lateral MOS-Capacitor (MOSCAP) RRM operating above 50GHz with modulation amplitude enhanced by a large plasma absorption within the MOS junction. A MOSCAP ring resonator modulator (RRM) model has been built using Lumerical software, in which the plasma effect is defined by adopting a reported superlinear rather than linear plasma absorption equation, which aligns well with our experimental results. The performance of the MOSCAP RRMs has been analyzed with different thicknesses of insulator oxide ( $ t_{\text{ox}}$ ). The modulation performance is enhanced with thinner $ t_{\text{ox}}$ down to 3 nm, giving a lower insertion loss and larger optical modulation amplitude (OMA) when benchmarked with a conventional depletion type RRM with a low \begin{equation*} V_{\pi }L \end{equation*} of 2.6-4.0 V $\cdot$ mm under a bias voltage $V_{\text{b}}$ 0-3V. High-speed operation of the MOSCAP RRM with radius 15 $\mu$ m demonstrated an average power insertion loss (IL $ _{\text{ave}}$ ) of 3.5 dB and one level insertion loss (IL $ _{\text{one}}$ ) of 2 dB for achieving a 3 dB dynamic ER at a data rate of 30 Gb/s and bit-error-rate (BER) \begin{equation*} < 1 \times 10^{-12} \end{equation*} . The same performance is possible at 50 Gb/s when feed-forward-equalization is enabled on the detection side. We also show the possibility of operating at 224 Gb/s using 4-level pulse amplitude modulation (PAM-4) for a MOSCAP RRM incorporating two active segments. The MOSCAP RRM provides an attractive solution to surpass the performance of the conventional depletion-type RRM, for which future performance scaling is limited with increased doping density towards $1 \times 10^{19} cm^{-3}$ .

3057 Background: 18 F-FDG PET/CT scans are used to assess tumor response for patients undergoing immunotherapy for metastatic melanoma (MM). Evaluating the relationship between features derived from quantitative medical image analysis and markers in blood samples such as circulating tumor DNA (ctDNA) may present an opportunity to discover imaging and blood biomarkers that can influence future patient care. This study evaluates the correlation of ctDNA and radiographic imaging features derived from FDG PET/CT. Methods: Whole-body FDG PET/CT scans from MM patients between 2014-2020 were retrospectively collected under IRB-approved protocol. Patients received pembrolizumab (n=20), ipilimumab (n=4), nivolumab (n=7), or a combination of ipilimumab and nivolumab (n=19). TRAQinform IQ software (AIQ Solutions) identified and quantified regions of interest suspicious of cancer (lesion-ROI), enabling extraction of imaging features including SUV max , SUV mean , and SUV total in baseline ( BL ) and follow-up ( FU ) images and change. Plasma ctDNA concentration (copies/ml) and frequency abundance (FA) were evaluated for the first sample ( 1 ) after immunotherapy had started (between baseline and first on-treatment scan), and the next available sample ( 2 ). Patients were further grouped by ctDNA mutation for analysis. Correlation between imaging features and ctDNA features was assessed using Spearman coefficient (ρ). Results: The patient cohort included 32 males and 19 females with average age of 62 years (range 23-83). For all 51 patients, moderate correlation was observed in SUV max,BL with ctDNA copies/ml 1 (ρ = 0.48, p<< 0.001) and ctDNA FA 1 (ρ = 0.49, p << 0.001) and in SUV total,BL with ctDNA FA 1 (ρ= 0.41, p < 0.001). In BRAF mutation patients (N=32), moderate correlations existed between SUV max,BL with ctDNA copies/ml 1 (ρ = 0.49, p = 0.0041 and ctDNA FA 1 (ρ = 0.05, p = 0.0035), and SUV total,BL with ctDNA copies/ml 2 (ρ = 0.42, p = 0.017), ctDNA FA 1 (ρ = 0.40, p = 0.023), and ctDNA FA 2 (ρ = 0.43, p = 0.012). All the other combinations displayed weak correlations (ρ < 0.39, p > 0.05). Conclusions: This study shows that in patients with MM receiving immunotherapy, quantitative features from blood biomarkers such as ctDNA correlate with FDG PET/CT imaging features. Combining blood biomarkers with imaging features that are spatially localizable may strengthen prognostication in this patient group. Further analysis is being performed with more types of ctDNA mutations.