An Online AI-Powered Interactive Histological Image Annotation Platform for Analyzing Intestinal Regenerating Crypts in Post-Irradiated Mice

PURPOSE/OBJECTIVE(S)

The goal of this project is to build an online AI-powered interactive annotation platform to accurately and efficiently annotate intestinal regenerating crypts in histological images of mice after abdominal irradiation.

MATERIALS/METHODS

The proposed platform is developed by the seamless integration of a front-end web client and a back-end server. Such client/server design allows the users to access the platform without software installation on local computers. Our front-end client is developed with SvelteJS + WebGL technology stack, allowing access from any common web browsers and enabling user interaction, such as image importing/visualization, interactive crypt annotating, and annotation saving/deleting. The back-end server is responsible for executing the tasks requested from the web client, for instance, image pre-processing, AI-based crypts automatic identification, and database management. The image preprocessing is designed to extract a single cross section image using morphological operations because multiple hematoxylin and eosin (H&E) stained jejunum cross sections from post-irradiated mice are scanned within one slide. The auto-crypt identification is powered by a trained and validated AI engine U-Net, classifying image grid tiles into two groups with and without regenerating crypts. The database is implemented with the self-contained SQLite to support recording and indexing the annotated grid tiles with regenerating crypts. The workflow for crypt analysis on this interactive platform has 5 steps: 1) manually import a whole H&E slide image; 2) auto-preprocess the slide by extracting single cross-section images; 3) auto-identify regenerating crypts with an AI engine; 4) interactively annotate (add, delete, modify) auto-identified crypt markers; 5) save and/or output the annotation to the database or the local drive.

RESULTS

The performance of the developed interactive crypt analysis platform was evaluated in aspects of accuracy and efficiency. The AI-powered crypt auto-identification accuracy was assessed by computing the mean absolute error (MAE) on crypt number per cross section between manual and auto annotation using a testing dataset containing 80 cross sections. It achieved an MAE of 3.5±4.8 crypts per cross section, and 81.25% of the cross sections have no more than 5 crypts difference. The efficiency was assessed under two conditions with the server on the cloud and a local computer. It took about 2-3 minutes to finish the entire workflow on the cloud, while 1-2 minutes on the local by saving ∼1 minute on image uploading.

CONCLUSION

The developed web client/server platform enables online automatic identification and interactive annotation of mice crypts in minutes. It is a convenient tool that allows accurate and efficient crypt analysis and can be extended for other histologic image analyses.

Deep Learning-Based Pipeline for Automatic Identification of Intestinal Regenerating Crypts in Mouse Histological Images

PURPOSE/OBJECTIVE(S)

A classical approach for evaluating normal tissue radiation response is to count the number of intestinal regenerating crypts in mouse histological images acquired after abdominal radiation. However, manual counting is time-consuming and subject to inter-observer variations. The goal of this study is to build a deep learning-based pipeline for automatically identifying intestinal regenerating crypts to facilitate high-throughput studies.

MATERIALS/METHODS

Sixty-six healthy C57BL/6 female mice underwent 16 MeV whole abdominal electron irradiation. The small bowel was collected from each mouse 4 days post-irradiation, and 9 jejunal cross-sections from each were processed together in a single slide. The slides were stained with hematoxylin and eosin (H&E) and subsequently scanned (x20), providing one electronic histological image per mouse. Regenerating crypts, consisting of more than 10 basophilic crypt epithelial cells, were manually identified using point annotations in histological images. The pipeline was built to take the input of the image containing 9 cross sections and automatically identify the regenerating crypts on each cross section. It mainly consists of two components, cross section segmentation using intensity thresholding and morphological operations and crypt identification using a UNet. The dataset was randomly split into 46, 10, and 10 slide images for UNet training, validation, and testing. Each slide image was split into grid tiles with a voxel size of 200 × 200, and 40 × 40 square masks were placed with centers at manual point annotations on tiles with regenerating crypts. 5203/5198 tiles (w/wo crypt mask) were extracted to train UNet by minimizing dice loss. The mask probability map generated by the UNet was post-processed to identify the crypt position. Postprocessing hyperparameters were tuned using the validation dataset. The model accuracy was evaluated using the testing dataset by computing the mean absolute error (MAE) of the crypt number averaged across all cross sections.

RESULTS

The number of regenerating crypts on testing cross sections ranges from 1 to 63. The testing cross-section-wise MAE achieved by the platform is 3.5±4.8 crypts. 81.25% of testing cross sections have absolute number differences less than or equal to 5 crypts.

CONCLUSION

Our established deep learning-based pipeline can accurately count the number of regenerating crypts in mouse intestinal histological images. We have integrated it into an online platform that enables automatic crypt identification and allows users to interactively modify auto-identified crypt annotations. The acquired annotations from the platform will be used to finetune the deep learning model to achieve better identification performance.

Anatomically Realistic 3D Printed Mouse Phantom for Multi-Institutional Benchmarking of FLASH and CONV Irradiation

PURPOSE/OBJECTIVE(S)

It is reported that about US$28B/year is spent on pre-clinical studies that are not reproducible. FLASH studies may suffer from the same reproducibility crisis due to the non-standard nature of the FLASH beamlines and the lack of dosimeters that can function at ultra-high dose-rates. There have been reports of different outcomes with regard to the FLASH effect across different institutions, even though similar beamlines, temporal structure, and nominal dose levels were used. This brings up the question of the accuracy of dosimetry under FLASH conditions for a fair comparison between FLASH and CONV. To answer this question, we develop and characterize an anatomically realistic 3D-printed mouse phantom to be used in a multi-institutional dosimetric benchmarking effort.

MATERIALS/METHODS

Mesh files for bony anatomy, lungs, and soft tissue derived from a CT scan of a mouse were converted to an editable 3D model. The 3D model was cut along the coronal plane and modified to allow the inclusion of radiographic film. A multi-material approach was employed to print the phantom. A dual-nozzle 3D printer was used, where one of the nozzles used Acrylonitrile butadiene styrene (ABS) to mimic soft tissue and the other nozzle used Polyactic acid (PLA) to mimic bone density. The two materials were used together in a single print. Lungs were approximated by lightweight PLA and were printed separately and inserted into corresponding cavities in the phantom. Hounsfield Units (HU) and print-to-print stability were verified. Radiographic films were laser cut for different anatomical sites. Two institutes took part in this study with data pending from 3 more institutions. The institutes were instructed to deliver 10 Gy to the plane of the film for the whole abdomen, whole lung, and brain irradiations. 2D dose maps were compared between FLASH and CONV, and the deviation from the prescribed dose was also measured.

RESULTS

The 3D-printed soft tissue, bone, and lung densities were measured to be ∼ 1.01 g/cc, 1.22 g/cc, and 0.44 g/cc, respectively. For soft tissue and bone, the Hounsfield unit (HU) difference from one print to another was < 10 HU. The greatest variation was within the lungs (54 HU), but this had a minimal effect on the dose distribution (<1%). For the two institutions that completed the survey, the maximum average difference between FLASH and CONV for all irradiations was 0.75 Gy (7.48%). The maximum average difference from the prescribed dose for all irradiations was 0.7 Gy (7.20%) across both institutions. The largest discrepancy was generally observed to be for lung irradiation, indicating that lack of treatment planning systems limits our ability to prescribe accurately in areas of inhomogeneities.

CONCLUSION

A 3D printed anatomically realistic mouse phantom was developed, characterized, and used in a multi-institutional dosimetric benchmarking effort. Such a study is paramount for the clinical translation of FLASH as it facilitates reduced variability from one institution to another.

Technical Infrastructure for Clinical Translation of Electron FLASH

PURPOSE/OBJECTIVE(S)

For safe clinical translation of electron FLASH, hardware tools for real-time beam control and software tools for treatment planning are necessary. The purpose of this study is to prototype high-throughput hardware for real-time beam control, along with accurate beam modeling of a modern clinical Linac configured to deliver FLASH dose-rates.

MATERIALS/METHODS

For real-time beam current monitoring, a beam current transformer (BCT) was initially coupled to a fast digitizer and its linearity was established by varying dose per pulse. The radiation pulse width was modified, and this change was measured using the BCT. The BCT was then used to measure the variability of dose per pulse and pulse width due to a mistuned linear accelerator system. Next, the BCT was interfaced with a field programmable gate array (FPGA) which provides the ability for high-throughput and deterministic control of the Linac based on dose accumulation. For beam modeling, the program, TOol for PArticle Simulation (TOPAS), was used to obtain beam parameters by using Bayesian optimization of the beam energy, source size, angular, and energy spread via comparison of simulated and representative dose profiles. The beam model would then be employed to calculate 3D dose distribution in a CT scan of a 3D-printed anatomically realistic mouse phantom.

RESULTS

The area under the current-time curve from the BCT exhibited excellent linearity (response = 12.80 nC/Gy) up to 2.5 Gy/Pulse (R2 = 0.99). The peak beam current for the electron FLASH beam was measured to be ∼10 mA for an instantaneous dose-rate of ∼5×105 Gy/s. The measured radiation pulse width agreed with the expected value (3.7 μs). The pulse width was then shortened and the measurement by the BCT indicated pulse widths of 1.8 μs and 0.5 μs corresponding to 0.7 Gy/pulse and 0.3 Gy/pulse, respectively. The beamline exhibited a ramp-up in dose per pulse and pulse width when using the automatic frequency controller (AFC). For the first pulse, the dose delivered was ∼0.1-0.3 Gy and the pulse width was 0.6 μs. The output stabilized to nominal values of dose and pulse width after 3-4 pulses. This ramp-up was mitigated by manually tuning the RF resonance with the AFC disabled, after which the BCT exhibited constant output and pulse width. The beam modeling work is in progress.

CONCLUSION

We demonstrated that a BCT can provide real-time measurement of per-pulse output suitable as input for FLASH beam control based on dose accumulation. The next steps are to quantify the accuracy of the dose control mechanism with the FPGA-based hardware. Potential failure modes will be identified and mitigated in parallel with the development of the hardware. A 3D-printed mouse phantom has been constructed to facilitate beam modeling work for treatment planning (in progress). On completion of this work, it is expected that we will have key infrastructure elements needed to move towards an eventual FDA investigational device exemption for clinical trials.

Hyperfractionated Reirradiation for Locally Recurrent Thoracic Tumors

PURPOSE/OBJECTIVE(S)

For patients with locally recurrent thoracic tumors or second primaries within previously irradiated volumes, hyperfractionated reirradiation (re-RT) may mitigate late toxicity compared to conventional fractionation, but clinical outcomes have not been extensively studied. We herein report our institutional experience with thoracic hyperfractionated reirradiation.

MATERIALS/METHODS

We identified 26 cases among 23 patients treated with re-RT to either primary or metastatic thoracic tumors, 60 Gy in 50 fractions, twice daily over 5 weeks using highly conformal image guided RT with motion management. Nineteen patients had dosimetry data available. The primary outcome was Grade (G2) or higher toxicity rates per CTCAEv5.0. Secondary endpoints were 12-month local control (LC), progression free survival (PFS)-determined by treating physician and/or multidisciplinary tumor board-and overall survival (OS).

RESULTS

Median follow-up was 13 months. Half had non-small cell lung cancer, 95.8% had ultracentral tumors, 57.7% had single prior thoracic RT course; 38.5%, 11.5% and 11.5% received concurrent chemotherapy, immunotherapy, and targeted agents, respectively. Minimum and median intervals between RT courses were 10 and 39.5 months, respectively; 94.7% of re-irradiation plans had overlapping 80% isodose volumes. Median OS and PFS were 13 and 10 months, respectively. Crude 12-month LC was 73.1%. Of those with a recurrence, the first recurrence occurred locally in 6 (54.6%), regionally in 3 (27.3%), and distantly in 8 (72.7%) patients. ≥G2 and ≥G3 toxicity rates were 30.8% and 7.69%, respectively (one G3 atrial fibrillation; one G5 pneumonitis). Using the American Radium Society guidelines for thoracic reirradiation, only 10.5% met all dose volume constraint recommendations.

CONCLUSION

Definitive hyperfractionated thoracic re-RT was well tolerated with promising local control. ≥G3 toxicities were rare. Patients should be counseled on the low but potential risk of life-threatening toxicity. Consensus guidelines for dose constraints may be difficult to meet in reirradiation setting; in this cohort, rates of severe toxicity were low despite exceeding putative constraints in most patients.

Predicting Local Control with Dosimetric Parameters in Patients Receiving Individualized Stereotactic Ablative Radiotherapy for Lung Tumors

PURPOSE/OBJECTIVE(S)

Stereotactic ablative radiotherapy (SABR) is an effective treatment option for lung tumors. The individualized lung tumor SABR (iSABR) trial was a phase II single-arm study that personalized lung tumor SABR dose and fractionation based on tumor size, location, and histology with very low rates of local recurrence (LR). A secondary analysis of this trial was conducted to assess for potential dosimetric predictors of LR, in order to help guide future clinical treatment planning.

MATERIALS/METHODS

From 2011 to 2018, local, regional and distant recurrence data were prospectively collected from 204 patients (261 lung SABR treatments) enrolled in a prospective trial. Baseline characteristics and treatment details were evaluated. Dosimetric and treatment plan parameters were evaluated for their potential to predict LR, using logistic regression and chi-squared analyses.

RESULTS

The majority of treated tumors were peripheral (71%, vs 29% central), primary lesions (76%, versus 24% metastatic), and of adenocarcinoma histology (67%, versus 13% squamous cell carcinoma and 19% other). The median follow-up was 24 months (range 2-95). Twenty-seven (10.3%) LRs occurred, with a median time to LR of 15 months (range 6-81 months). There were no significant associations between the overall cohort and the dosimetric parameters. However, for the multi-fraction cohort, an increased proportion of the PTV receiving 110% and 115% of the prescription dose were associated with lower LR (p = 0.01 and p = 0.01 respectively). Specifically for the 50 Gy in 4 fraction cohort, an increased D1cc, D0.03cc, as well as the proportion of the PTV receiving 110%, 115%, and 120% of the prescription dose were associated with lower LR (p < 0.001, p = 0.001, p = 0.003, p < 0.001, p = 0.004, respectively). There was no association of LR with prescription dose expressed as biologically effective dose using an alpha/beta of 10 Gy (BED10), D99%, or single- versus multi-fraction regimens.

CONCLUSION

SABR for lung tumors using the individualized protocol on this trial showed excellent LR rates. We identified dosimetric parameters that were associated with LR, including V110% and V115% within the multi-fraction cohort, as well as the 50 Gy in 4 fraction cohort the D1cc, D0.03cc, and proportions of the PTV receiving 110%, 115%, and 120% of the prescription dose in the 50 Gy in 4 fraction cohort. Optimal thresholds for these parameters will be identified in further analyses. There did not appear to be an association with LR and BED10, D99%, or comparing single- vs multi-fraction regimens.

Investigating Dosimetry and Imaging Biomarkers for Prediction of Major Adverse Cardiac Events Following Locally Advanced Non-Small Cell Lung Cancer Radiotherapy

PURPOSE/OBJECTIVE(S)

Thoracic radiotherapy (RT) may confer major adverse cardiac events (MACE) following treatment. Mean heart dose positively associates with MACE and recent studies show cardiac substructure dosimetry improves MACE prediction. Use of imaging biomarkers with cardiac substructure dose has not been studied for prediction of MACE. We sought to develop an integrated model for cardiac substructure dose and baseline coronary artery calcium (CAC) scoring and establish its relationship to MACE.

MATERIALS/METHODS

A retrospective cohort analysis was performed of consecutive patients with locally advanced non-small cell lung cancer (NSCLC) treated with definitive RT from 2006-2018 at a single institution. Demographics, medical history, cardiac events, and treatments received were recorded. Cardiac substructures were contoured, including the left descending artery (LAD), left main coronary artery (LMCA), left circumflex (LCX), right coronary artery (RCA), TotalLeft (LAD+LMCA+LCX), and TotalCor (TotalLeft+RCA). Doses were measured in 2 Gy equivalent dose. CAC was scored by visual assessment and compared to established automated Agatston scoring. Primary endpoint was MACE incidence. Receiver operating characteristic (ROC) curves assessed dose and CAC metric model performance. Threshold modeling was conducted using the log rank statistic with 95% confidence intervals measured using bootstrap resampling with 1000 iterations. Competing risk models adjusted for death were used to measure cumulative incidence of MACE as well as in univariable and multivariable risk regression modeling. Pearson correlations were used to validate CAC scoring. P-values were two tailed and considered significant at P≤0.05.

RESULTS

Of 233 eligible patients, 61.4% were male with a 68.1 years (range 34.9-90.7) median age. Median follow-up was 73.7 months (range 1.6-153.9). Median overall survival was 34.8 months. Following RT, 22.3% experienced at least one cardiac event at a median time of 21.5 months (range 1.7-118.9). Visual CAC scoring showed significant correlation with automated Agatston scoring (r = 0.72, P=1e-5). While left sided coronary arteries (TotalLeft), mean heart dose (MHD) and CAC scores individually predicted for MACE (AUC = 0.56-0.59), a multivariable model of TotalLeft CAC had the highest ROC analysis performance (AUC = 0.69). On univariable and multivariable competing risk regression analyses, TotalLeft V15 Gy >2.53 cc and CAC score >5 independently associated with MACE (P<0.05). A model incorporating age, TotalLeft CAC>5 and V15>2.53cc, showed incrementally higher MACE incidences for low (9.3%), intermediate (18.4%), and high-risk groups (27.7%) (P<0.01).

CONCLUSION

RT-induced MACE occurs in >20% of those undergoing thoracic RT in a median time of <2 years. We validate significant associations between TotalLeft RT dose and MACE and establish CAC as a predictive risk factor. These findings may serve to inform personalized RT and future cardiac risk in locally advanced NSCLC.

Esophageal tolerance to high-dose stereotactic ablative radiotherapy

Dose-volume parameters are needed to guide the safe administration of stereotactic ablative radiotherapy (SABR). We report on esophageal tolerance to high-dose hypofractionated radiation in patients treated with SABR. Thirty-one patients with spine or lung tumors received single- or multiple-fraction SABR to targets less than 1 cm from the esophagus. End points evaluated include D(5cc) (minimum dose in Gy to 5 cm(3) of the esophagus receiving the highest dose), D(2cc) , D(1cc) , and D(max) (maximum dose to 0.01 cm(3) ). Multiple-fraction treatments were correlated using the linear quadratic and linear quadratic-linear/universal survival models. Three esophageal toxicity events occurred, including esophagitis (grade 2), tracheoesophageal fistula (grade 4-5), and esophageal perforation (grade 4-5). Chemotherapy was a cofactor in the high-grade events. The median time to development of esophageal toxicity was 4.1 months (range 0.6-6.1 months). Two of the three events occurred below a published D(5cc) threshold, all three were below a D(2cc) threshold, and one was below a D(max) threshold. We report a dosimetric analysis of incidental dose to the esophagus from SABR. High-dose hypofractionated radiotherapy led to a number of high-grade esophageal adverse events, suggesting that conservative parameters to protect the esophagus are necessary when SABR is used, especially in the setting of chemotherapy or prior radiotherapy.

Initial evaluation of 18F-fluorothymidine (FLT) PET/CT scanning for primary pancreatic cancer

PURPOSE

The aim of this study was to evaluate the potential of (18)F-fluorothymidine (FLT) PET/CT for imaging pancreatic adenocarcinoma.

METHODS

This was a pilot study of five patients (four males, one female) with newly diagnosed and previously untreated pancreatic adenocarcinoma. Patients underwent FLT PET/CT, (18)F-fluorodeoxyglucose (FDG) PET/CT, and contrast-enhanced CT scanning before treatment. The presence of cancer was confirmed by histopathological analysis at the time of scanning in all five patients. The degree of FLT and FDG uptake at the primary tumor site was assessed using visual interpretation and semi-quantitative SUV analyses.

RESULTS

The primary tumor size ranged from 2.5 x 2.8 cm to 3.5 x 7.0 cm. The SUV of FLT uptake within the primary tumor ranged from 2.1 to 3.1. Using visual interpretation, the primary cancer could be detected from background activity in two of five patients (40%) on FLT PET/CT. By comparison, FDG uptake was higher in each patient with a SUV range of 3.4 to 10.8, and the primary cancer could be detected from background in all five patients (100%).

CONCLUSIONS

In this pilot study of five patients with primary pancreatic adenocarcinoma, FLT PET/CT scanning showed poor lesion detectability and relatively low levels of radiotracer uptake in the primary tumor.

Plasmablastic lymphoma presenting in a human immunodeficiency virus-negative patient: a case report

Plasmablastic lymphoma (PBL), an aggressive non-Hodgkin's lymphoma that carries a poor prognosis, previously has been identified almost exclusively in patients infected with the human immunodeficiency virus (HIV). We present a case of a 42-year-old HIV-negative patient presenting with an isolated nasal cavity mass, the typical presentation for PBL. The patient was given systemic chemotherapy, central nervous system prophylaxis, and consolidative locoregional radiotherapy and achieved a complete clinical response. This case suggests PBL should be considered in HIV-negative patients with characteristic findings.

A new sample preparation method for biological soft X-ray microscopy: nitrogen-based contrast and radiation tolerance properties of glycol methacrylate-embedded and sectioned tissue

We describe the preparation of a biological tissue for imaging in a transmission soft X-ray microscope. Sections of exocrine pancreas embedded in glycol methacrylate polymer, an embedding medium widely used in visible light and electron microscopy, were examined. Contrast was based primarily on the nitrogen content of the tissue, and dual-wavelength imaging at the nitrogen K-shell absorption edge was used to map the distribution and provide quantitative densitometry of both the protein and embedding matrix components of the sample. The measurements were calibrated by obtaining the absorption spectrum of protein near the nitrogen edge. The contrast was consistent and reproducible, making possible the first large-scale X-ray microscopic study on sections of plastic-embedded soft tissue. At radiation doses of up to 10(8) Gray, much more than required for routine imaging, no distortion and little mass loss were observed. This sample preparation method should permit routine imaging of tissues in X-ray microscopes, previously a difficult task, as well as multimodal imaging (using visible light, X-ray, electron, and scanned probe microscopies) on the same sample.

Automatic image acquisition, calibration and montage assembly for biological X-ray microscopy

We describe a system for the automatic acquisition and processing of digital images in a high-resolution X-ray microscope, including the formation of large-field high-resolution image montages. A computer-controlled sample positioning stage provides approximate coordinates for each high-resolution subimage. Individual subimages are corrected to compensate for time-varying, non-uniform illumination and CCD-related artefacts. They are then automatically assembled into a montage. The montage assembly algorithm is designed to use the overlap between each subimage and multiple neighbours to improve the performance of the registration step and the fidelity of the result. This is accomplished by explicit use of recorded stage positions, optimized ordering of subimage insertion, and registration of subimages to the developing montage. Using this procedure registration errors are below the resolution limit of the microscope (43 nm). The image produced is a seamless, large-field montage at full resolution, assembled automatically without human intervention. Beyond this, it is also an accurate X-ray transmission map that allows the quantitative measurement of anatomical and chemical features of the sample. Applying these tools to a biological problem, we have conducted the largest X-ray microscopical study to date.