Dose and fractionation regimen for brachytherapy boost in cervical cancer in the US.

PURPOSE: Curative-intent radiotherapy for locally advanced and select early stage cervical cancer in the US includes external beam radiotherapy (EBRT) with brachytherapy. Although there are guidelines for brachytherapy dose and fractionation regimens, there are limited data on practice patterns. This study aims to evaluate the contemporary utilization of cervical cancer brachytherapy in the US and its association with patient demographics and facility characteristics. METHODS: We retrospectively analyzed clinical covariates of cervical cancer patients diagnosed and treated in 2018-2020 with curative-intent radiotherapy from the 2020 National Cancer Database. Associations between patient and institutional factors with the number of brachytherapy fractions were identified with logistic regression. Factors with association (p < 0.10) were then included in a multivariable logistic regression model. All tests were two-sided with significance <0.05 unless specified otherwise. RESULTS: Among the eligible 2517 patients, 97.3% received HDR or LDR and is further analyzed. More patients received HDR than LDR brachytherapy (98.9% vs 1.1%) and intracavitary than interstitial brachytherapy (86.4% vs 13.6%). The most common number of HDR fractions prescribed were 5 (51.0%), 4 (32.9%), and 3 (8.6%). After adjusting for the other variables in the model, ethnicity, private insurance status, overall insurance status, and facility type were the only factors that were significantly associated with the number of brachytherapy factions (p < 0.0001, p = 0.028, p = 0.001, and p < 0.0001, respectively, n = 2184). CONCLUSIONS: In the US, various HDR brachytherapy regimens are utilized depending on patient and institutional factors. Future research may optimize cervical cancer brachytherapy by correlating specific dose and fractionation regimens with patient outcomes.

Using Machine Learning to Predict Response to Image-guided Therapies for Hepatocellular Carcinoma.

Interventional oncology is a rapidly growing field with advances in minimally invasive image-guided local-regional treatments for hepatocellular carcinoma (HCC), including transarterial chemoembolization, transarterial radioembolization, and thermal ablation. However, current standardized clinical staging systems for HCC are limited in their ability to optimize patient selection for treatment as they rely primarily on serum markers and radiologist-defined imaging features. Given the variation in treatment responses, an updated scoring system that includes multidimensional aspects of the disease, including quantitative imaging features, serum markers, and functional biomarkers, is needed to optimally triage patients. With the vast amounts of numerical medical record data and imaging features, researchers have turned to image-based methods, such as radiomics and artificial intelligence (AI), to automatically extract and process multidimensional data from images. The synthesis of these data can provide clinically relevant results to guide personalized treatment plans and optimize resource utilization. Machine learning (ML) is a branch of AI in which a model learns from training data and makes effective predictions by teaching itself. This review article outlines the basics of ML and provides a comprehensive overview of its potential value in the prediction of treatment response in patients with HCC after minimally invasive image-guided therapy.

COVID-19 mortality prediction in the intensive care unit with deep learning based on longitudinal chest X-rays and clinical data.

OBJECTIVES: We aimed to develop deep learning models using longitudinal chest X-rays (CXRs) and clinical data to predict in-hospital mortality of COVID-19 patients in the intensive care unit (ICU). METHODS: Six hundred fifty-four patients (212 deceased, 442 alive, 5645 total CXRs) were identified across two institutions. Imaging and clinical data from one institution were used to train five longitudinal transformer-based networks applying five-fold cross-validation. The models were tested on data from the other institution, and pairwise comparisons were used to determine the best-performing models. RESULTS: A higher proportion of deceased patients had elevated white blood cell count, decreased absolute lymphocyte count, elevated creatine concentration, and incidence of cardiovascular and chronic kidney disease. A model based on pre-ICU CXRs achieved an AUC of 0.632 and an accuracy of 0.593, and a model based on ICU CXRs achieved an AUC of 0.697 and an accuracy of 0.657. A model based on all longitudinal CXRs (both pre-ICU and ICU) achieved an AUC of 0.702 and an accuracy of 0.694. A model based on clinical data alone achieved an AUC of 0.653 and an accuracy of 0.657. The addition of longitudinal imaging to clinical data in a combined model significantly improved performance, reaching an AUC of 0.727 (p = 0.039) and an accuracy of 0.732. CONCLUSIONS: The addition of longitudinal CXRs to clinical data significantly improves mortality prediction with deep learning for COVID-19 patients in the ICU. KEY POINTS: • Deep learning was used to predict mortality in COVID-19 ICU patients. • Serial radiographs and clinical data were used. • The models could inform clinical decision-making and resource allocation.

DICOM Image ANalysis and Archive (DIANA): an Open-Source System for Clinical AI Applications.

In the era of data-driven medicine, rapid access and accurate interpretation of medical images are becoming increasingly important. The DICOM Image ANalysis and Archive (DIANA) system is an open-source, lightweight, and scalable Python interface that enables users to interact with hospital Picture Archiving and Communications Systems (PACS) to access such data. In this work, DIANA functionality was detailed and evaluated in the context of retrospective PACS data retrieval and two prospective clinical artificial intelligence (AI) pipelines: bone age (BA) estimation and intra-cranial hemorrhage (ICH) detection. DIANA orchestrates activity beginning with post-acquisition study discovery and ending with online notifications of findings. For AI applications, system latency (exam completion to system report time) was quantified and compared to that of clinicians (exam completion to initial report creation time). Mean DIANA latency was 9.04 ± 3.83 and 20.17 ± 10.16 min compared to clinician latency of 51.52 ± 58.9 and 65.62 ± 110.39 min for BA and ICH, respectively, with DIANA latencies being significantly lower (p < 0.001). DIANA's capabilities were also explored and found effective in retrieving and anonymizing protected health information for "big-data" medical imaging research and analysis. Mean per-image retrieval times were 1.12 ± 0.50 and 0.08 ± 0.01 s across x-ray and computed tomography studies, respectively. The data herein demonstrate that DIANA can flexibly integrate into existing hospital infrastructure and improve the process by which researchers/clinicians access imaging repository data. This results in a simplified workflow for large data retrieval and clinical integration of AI models.

Gender Representation Among Academic US Radiation Oncology Department Chairs: Who are They?

Purpose: Women are underrepresented in academic radiation oncology (RO), particularly in leadership positions. In this study, we sought to better understand the characteristics of individuals who currently serve as academic RO chairpersons at institutions with an associated Accreditation Council for Graduate Medical Education-accredited RO residency training program. Methods and Materials: We created a database of academic RO chairpersons in the United States by using publicly available sources, including residency training program websites, hospital/institutional websites, Doximity, LinkedIn, the American Society for Radiation Oncology (ASTRO) website, the American College of Radiation Oncology website, and the National Plan and Provider Enumeration System National Provider Identifier Registry. We used the χ2 Goodness of Fit test, Mann-Whitney U test, and Fisher exact test via R version 4.1.1 to evaluate for statistical significance among categorical variables, medians, and proportions, respectively. Results: We identified 85 of the 90 chairpersons (94.4%) currently serving at institutions with an Accreditation Council for Graduate Medical Education-accredited RO residency training program, 5 of whom hold interim positions and were thus excluded from further analyses. Of the remaining 80 chairpersons, 9 (11.3%) are women, and 71 (88.8%) are men (P < .01). Seventy-six chairpersons (95.0%) are full professors, and 19 (23.8%) hold dual MD PhD degrees. Thirty-two chairpersons (40.0%) hold an official leadership role in a cancer center affiliated with their current institution (43.7% of men vs 11.1% of women; P = .08). Seventy-three chairpersons (91.3%) secured their current positions a median of 16 years (range, 6-33 years) after completing RO residency. Thirty-five chairpersons (43.8%) were promoted to chair from positions within their current institutions (40.8% of men vs 66.7% of women; P = .17). The majority of chairpersons are ASTRO Fellows (62.5%); notably fewer are ASTRO (5.0%) or American College of Radiation Oncology (2.5%) Gold Medalists. Eight RO residency programs trained more than half of current chairpersons. Conclusion: Significantly more men than women currently serve as RO chairpersons. Future interventions that promote the recruitment, retention, and promotion of talented women in academic RO should be considered.

Adjuvant Reirradiation With Proton Therapy in Head and Neck Squamous Cell Carcinoma.

Purpose: For patients with head and neck squamous cell carcinoma (HNSCC), locoregional failure and second primary tumors are common indications for adjuvant reirradiation (re-RT). Given an absence of clear consensus on the role of adjuvant re-RT, we sought to assess histopathologic risk factors of patients with HNSCC and their resulting outcomes after adjuvant re-RT with proton therapy. Methods and Materials: We conducted a retrospective analysis of patients with HNSCC who underwent salvage surgery at our institution followed by adjuvant re-RT with proton therapy over 1.5 years. All included patients received prior radiation therapy. The Kaplan-Meier method was used to evaluate locoregional recurrence-free survival and overall survival. Results: The cohort included 22 patients, with disease subsites, including oropharynx, oral cavity, hypopharynx, larynx, and nasopharynx. Depending on adverse pathologic features, adjuvant re-RT to 66 Gy (32% of cohort) or 60 Gy (68%), with (59%) or without (41%) concurrent systemic therapy was administered. The majority (86%) completed re-RT with no reported treatment delay; 3 patients experienced grade ≥3 acute Common Terminology Criteria for Adverse Events toxicity and no patient required enteral feeding tube placement during re-RT. Median follow-up was 21.0 months (IQR, 11.7-25.2 months). Five patients had biopsy-proven disease recurrences a median of 5.9 months (IQR, 3.8-9.7 months) after re-RT. Locoregional recurrence-free survival was 95.2%, 70.2%, 64.8% at 6, 12, and 24 months, respectively. OS was 100%, 79.2%, and 79.2% at 6, 12, and 24 months, respectively. Four patients had osteoradionecrosis on imaging a median of 13.2 months (IQR, 8.7-17.4 months) after re-RT, with 2 requiring surgical intervention. Conclusions: Adjuvant re-RT for patients with HNSCC was well-tolerated and offered reasonable local control in this high-risk cohort but appears to be associated with a risk of osteoradionecrosis. Additional study and longer follow-up could help define optimal patient management in this patient population.

An automated COVID-19 triage pipeline using artificial intelligence based on chest radiographs and clinical data.

While COVID-19 diagnosis and prognosis artificial intelligence models exist, very few can be implemented for practical use given their high risk of bias. We aimed to develop a diagnosis model that addresses notable shortcomings of prior studies, integrating it into a fully automated triage pipeline that examines chest radiographs for the presence, severity, and progression of COVID-19 pneumonia. Scans were collected using the DICOM Image Analysis and Archive, a system that communicates with a hospital's image repository. The authors collected over 6,500 non-public chest X-rays comprising diverse COVID-19 severities, along with radiology reports and RT-PCR data. The authors provisioned one internally held-out and two external test sets to assess model generalizability and compare performance to traditional radiologist interpretation. The pipeline was evaluated on a prospective cohort of 80 radiographs, reporting a 95% diagnostic accuracy. The study mitigates bias in AI model development and demonstrates the value of an end-to-end COVID-19 triage platform.