Identifying bias in models that detect vocal fold paralysis from audio recordings using explainable machine learning and clinician ratings.

Detecting voice disorders from voice recordings could allow for frequent, remote, and low-cost screening before costly clinical visits and a more invasive laryngoscopy examination. Our goals were to detect unilateral vocal fold paralysis (UVFP) from voice recordings using machine learning, to identify which acoustic variables were important for prediction to increase trust, and to determine model performance relative to clinician performance. Patients with confirmed UVFP through endoscopic examination (N = 77) and controls with normal voices matched for age and sex (N = 77) were included. Voice samples were elicited by reading the Rainbow Passage and sustaining phonation of the vowel "a". Four machine learning models of differing complexity were used. SHapley Additive exPlanations (SHAP) was used to identify important features. The highest median bootstrapped ROC AUC score was 0.87 and beat clinician's performance (range: 0.74-0.81) based on the recordings. Recording durations were different between UVFP recordings and controls due to how that data was originally processed when storing, which we can show can classify both groups. And counterintuitively, many UVFP recordings had higher intensity than controls, when UVFP patients tend to have weaker voices, revealing a dataset-specific bias which we mitigate in an additional analysis. We demonstrate that recording biases in audio duration and intensity created dataset-specific differences between patients and controls, which models used to improve classification. Furthermore, clinician's ratings provide further evidence that patients were over-projecting their voices and being recorded at a higher amplitude signal than controls. Interestingly, after matching audio duration and removing variables associated with intensity in order to mitigate the biases, the models were able to achieve a similar high performance. We provide a set of recommendations to avoid bias when building and evaluating machine learning models for screening in laryngology.

Enhanced recovery program in ambulatory pediatric urology: A quality improvement initiative.

BACKGROUND: Enhanced Recovery After Surgery (ERAS) was established in 2001 for adult patients undergoing complex procedures. ERAS in adult ambulatory surgery later followed with similar positive outcomes. For the pediatric population, ERAS implementation has shown promising results in complex surgeries such as bladder reconstruction. Its application in pediatric ambulatory surgery has only recently been reported. We hereby report a Quality Improvement initiative in implementing an Enhanced Recovery Protocol (ERP) for pediatric urology in an ambulatory surgery center. METHODS: A project was launched to evaluate and implement enhanced recovery elements into an institutional Enhanced Recovery Protocol (ERP). These included reliance on peripheral nerve blocks for all inguinal and genital cases and reduction of opioids intraoperatively and postoperatively. Improvements were placed into a project plan broken into one preparation phase to collect baseline data and three implementation phases to enhance existing and implement new elements. The implementation phase went through iterative Plan-Do-Study-Act (PDSA) cycles for all sub-projects. Team countermeasures were based on available evidence. A consensus process was used to resolve disagreement. Monthly meetings were held to share real-time data, gather new feedback, and modify plans as needed. The primary outcome measures selected were percent intraoperative opioid use, percent opioid prescribing, mean PACU length of stay, and average number of opioid doses prescribed. Secondary outcome measures were mean maximum pain score in PACU, PACU rescue rate for PONV, and patient/family satisfaction scores. Post-implementation data for 18 months was included for evaluation. Statistical process control methodology was used. RESULTS: The total number of participants was 3306: 561 (baseline), 220 (Phase 1) 356 (Phase 2) and 527 (Phase 3), 1642 (post-implementation). Intraoperative opioid use was eliminated in >99% of cases. Post-operative opioid prescribing was reduced from 30% to 15% of patients. The number of opioid doses was also reduced from an average of 7.6 to 6.1 doses. There was no change for the mean maximum pain score in the recovery room despite elimination of opioids. Patient/family satisfaction scores were high and sustained throughout the period of study (9.8/10). Balancing measures such as return to the operating room within 30 days and return to the emergency department within 7 days were unchanged. CONCLUSIONS: This QI project demonstrated the feasibility of a pediatric enhanced recovery protocol in a urology ambulatory surgery setting. With implementation of this protocol, intraoperative opioid use was virtually eliminated, and opioid prescribing was reduced without affecting pain scores or post-operative complications.

Transforming into a Learning Health System: A Quality Improvement Initiative.

Background: The Institute of Medicine introduced the Learning Healthcare System concept in 2006. The system emphasizes quality, safety, and value to improve patient outcomes. The Bellevue Clinic and Surgical Center is an ambulatory surgical center that embraces continuous quality improvement to provide exceptional patient-centered care to the pediatric surgical population. Methods: We used statistical process control charts to study the hospital's electronic health record data. Over the past 7 years, we have focused on the following areas: efficiency (surgical block time use), effectiveness (providing adequate analgesia after transitioning to an opioid-sparing protocol), efficacy (creating a pediatric enhanced recovery program), equity (evaluating for racial disparities in surgical readmission rates), and finally, environmental safety (tracking and reducing our facility's greenhouse gas emissions from inhaled anesthetics). Results: We have seen improvement in urology surgery efficiency, resulting in a 37% increase in monthly surgical volume, continued adaptation to our opioid-sparing protocol to further reduce postanesthesia care unit opioid administration for tonsillectomy and adenoidectomy cases, successful implementation of an enhanced recovery program, continued work to ensure equitable healthcare for our patients, and more than 85% reduction in our facility's greenhouse gas emissions from inhaled anesthetics. Conclusions: The Bellevue Clinic and Surgical Center facility is a living example of a learning health system, which has evolved over the years through continued patient-centered QI work. Our areas of emphasis, including efficiency, effectiveness, efficacy, equity, and environmental safety, will continue to impact the community we serve positively.

Long-Term Analysis of NRG Oncology RTOG 0415: A Randomized Phase III Noninferiority Study Comparing Two Fractionation Schedules in Patients With Low-Risk Prostate Cancer.

Clinical trials frequently include multiple end points that mature at different times. The initial report, typically based on the primary end point, may be published when key planned co-primary or secondary analyses are not yet available. Clinical Trial Updates provide an opportunity to disseminate additional results from studies, published in JCO or elsewhere, for which the primary end point has already been reported.NRG Oncology RTOG 0415 is a randomized phase III noninferiority (NI) clinical trial comparing conventional fractionation (73.8 Gy in 41 fractions) radiotherapy (C-RT) with hypofractionation (H-RT; 70 Gy in 28) in patients with low-risk prostate cancer. The study included 1,092 protocol-eligible patients initially reported in 2016 with a median follow-up of 5.8 years. Updated results with median follow-up of 12.8 years are now presented. The estimated 12-year disease-free survival (DFS) is 56.1% (95% CI, 51.5 to 60.5) for C-RT and 61.8% (95% CI, 57.2 to 66.0) for H-RT. The DFS hazard ratio (H-RT/C-RT) is 0.85 (95% CI, 0.71 to 1.03), confirming NI (P < .001). Twelve-year cumulative incidence of biochemical failure (BF) was 17.0% (95% CI, 13.8 to 20.5) for C-RT and 9.9% (95% CI, 7.5 to 12.6) for H-RT. The HR (H-RT/C-RT) comparing biochemical recurrence between the two arms was 0.55 (95% CI, 0.39 to 0.78). Late grade ≥3 GI adverse event (AE) incidence is 3.2% (C-RT) versus 4.4% (H-RT), with relative risk (RR) for H-RT versus C-RT 1.39 (95% CI, 0.75 to 2.55). Late grade ≥3 genitourinary (GU) AE incidence is 3.4% (C-RT) versus 4.2% (H-RT), RR 1.26 (95% CI, 0.69 to 2.30). Long-term DFS is noninferior with H-RT compared with C-RT. BF is less with H-RT. No significant differences in late grade ≥3 GI/GU AEs were observed between assignments (ClinicalTrials.gov identifier: NCT00331773).

Investigating the weekend effect in decompressive thoracolumbar hemilaminectomy for acute intervertebral disc extrusion: An observational cohort study of 460 cases (2018-2023).

OBJECTIVE: To investigate if weekend surgery is associated with poorer outcomes in dogs with acute thoracolumbar intervertebral disc extrusion (IVDE) undergoing decompressive thoracolumbar hemilaminectomy. STUDY DESIGN: Retrospective observational cohort study. SAMPLE POPULATION: A total of 460 consecutive cases were reviewed, with 401 dogs undergoing weekday surgery (Cohort WD), and 59 dogs undergoing weekend surgery (Cohort WE). METHODS: Medical records of a surgical referral center in the UK were reviewed. Preoperative patient demographic and clinical data, and postoperative outcome data were collected with a minimum 28-day follow-up period. Multivariable logistic regression analysis was used to model the odds of a negative outcome. RESULTS: Cohort WE had a higher preoperative proportion of nonambulatory dogs (p = .0115) but there were no significant differences between the nonambulatory (p = .3762) and deep-pain negative subgroups (p = .6199). Cohort WE had a higher risk of not recovering ambulation compared to Cohort WD [79.2% vs. 91.6% recovery; adjusted OR 3.010 (95% CI: 1.259-7.190); p = .0132] and had a higher risk of postoperative morbidity [32.2% vs. 17.2%; adjusted OR 2.015 (95% CI: 1.089-3.729); p = .0257]. There were no significant differences in other outcome measures between cohorts. CONCLUSION: Weekend surgery in canine decompressive thoracolumbar hemilaminectomy may be associated with poorer patient outcomes, specifically higher postoperative morbidity and a poorer rate of recovery of ambulation. CLINICAL SIGNIFICANCE: This study demonstrates a weekend effect in veterinary surgery, which may be important in surgical decision-making in acute thoracolumbar IVDE. Further scrutiny of the patient's journey through the veterinary healthcare system is warranted.

Intra-fractional geometric and dose/volume metric variations of magnetic resonance imaging-guided stereotactic radiotherapy of prostate bed after radical prostatectomy.

Background and purpose: Magnetic Resonance Imaging (MRI)-guided Stereotactic body radiotherapy (SBRT) treatment to prostate bed after radical prostatectomy has garnered growing interests. The aim of this study is to evaluate intra-fractional anatomic and dose/volume metric variations for patients receiving this treatment. Materials and methods: Nineteen patients who received 30-34 Gy in 5 fractions on a 0.35T MR-Linac were included. Pre- and post-treatment MRIs were acquired for each fraction (total of 75 fractions). The Clinical Target Volume (CTV), bladder, rectum, and rectal wall were contoured on all images. Volumetric changes, Hausdorff distance, Mean Distance to Agreement (MDA), and Dice similarity coefficient (DSC) for each structure were calculated. Median value and Interquartile range (IQR) were recorded. Changes in target coverage and Organ at Risk (OAR) constraints were compared and evaluated using Wilcoxon rank sum tests at a significant level of 0.05. Results: Bladder had the largest volumetric changes, with a median volume increase of 48.9 % (IQR 28.9-76.8 %) and a median MDA of 5.1 mm (IQR 3.4-7.1 mm). Intra-fractional CTV volume remained stable with a median volume change of 1.2 % (0.0-4.8 %). DSC was 0.97 (IQR 0.94-0.99). For the dose/volume metrics, there were no statistically significant changes observed except for an increase in bladder hotspot and a decrease of bladder V32.5 Gy and mean dose. The CTV V95% changed from 99.9 % (IQR 98.8-100 %) to 99.6 % (IQR 93.9-100 %). Conclusion: Despite intra-fractional variations of OARs, CTV coverage remained stable during MRI-guided SBRT treatments for the prostate bed.

Motion compensated cone-beam CT reconstruction using ana priorimotion model from CT simulation: a pilot study.

Objective. To combat the motion artifacts present in traditional 4D-CBCT reconstruction, an iterative technique known as the motion-compensated simultaneous algebraic reconstruction technique (MC-SART) was previously developed. MC-SART employs a 4D-CBCT reconstruction to obtain an initial model, which suffers from a lack of sufficient projections in each bin. The purpose of this study is to demonstrate the feasibility of introducing a motion model acquired during CT simulation to MC-SART, coined model-based CBCT (MB-CBCT).Approach. For each of 5 patients, we acquired 5DCTs during simulation and pre-treatment CBCTs with a simultaneous breathing surrogate. We cross-calibrated the 5DCT and CBCT breathing waveforms by matching the diaphragms and employed the 5DCT motion model parameters for MC-SART. We introduced the Amplitude Reassignment Motion Modeling technique, which measures the ability of the model to control diaphragm sharpness by reassigning projection amplitudes with varying resolution. We evaluated the sharpness of tumors and compared them between MB-CBCT and 4D-CBCT. We quantified sharpness by fitting an error function across anatomical boundaries. Furthermore, we compared our MB-CBCT approach to the traditional MC-SART approach. We evaluated MB-CBCT's robustness over time by reconstructing multiple fractions for each patient and measuring consistency in tumor centroid locations between 4D-CBCT and MB-CBCT.Main results. We found that the diaphragm sharpness rose consistently with increasing amplitude resolution for 4/5 patients. We observed consistently high image quality across multiple fractions, and observed stable tumor centroids with an average 0.74 ± 0.31 mm difference between the 4D-CBCT and MB-CBCT. Overall, vast improvements over 3D-CBCT and 4D-CBCT were demonstrated by our MB-CBCT technique in terms of both diaphragm sharpness and overall image quality.Significance. This work is an important extension of the MC-SART technique. We demonstrated the ability ofa priori5DCT models to provide motion compensation for CBCT reconstruction. We showed improvements in image quality over both 4D-CBCT and the traditional MC-SART approach.

Results of an Artificial Intelligence-Based Image Review System to Detect Patient Misalignment Errors in a Multi-institutional Database of Cone Beam Computed Tomography-Guided Radiation Therapy.

PURPOSE: Present knowledge of patient setup and alignment errors in image guided radiation therapy (IGRT) relies on voluntary reporting, which is thought to underestimate error frequencies. A manual retrospective patient-setup misalignment error search is infeasible owing to the bulk of cases to be reviewed. We applied a deep learning-based misalignment error detection algorithm (EDA) to perform a fully automated retrospective error search of clinical IGRT databases and determine an absolute gross patient misalignment error rate. METHODS AND MATERIALS: The EDA was developed to analyze the registration between planning scans and pretreatment cone beam computed tomography scans, outputting a misalignment score ranging from 0 (most unlikely) to 1 (most likely). The algorithm was trained using simulated translational errors on a data set obtained from 680 patients treated at 2 radiation therapy clinics between 2017 and 2022. A receiver operating characteristic analysis was performed to obtain target thresholds. DICOM Query and Retrieval software was integrated with the EDA to interact with the clinical database and fully automate data retrieval and analysis during a retrospective error search from 2016 to 2017 and from 2021 to 2022 for the 2 institutions, respectively. Registrations were flagged for human review using both a hard-thresholding method and a prediction trending analysis over each individual patient's treatment course. Flagged registrations were manually reviewed and categorized as errors (>1 cm misalignment at the target) or nonerrors. RESULTS: A total of 17,612 registrations were analyzed by the EDA, resulting in 7.7% flagged events. Three previously reported errors were successfully flagged by the EDA, and 4 previously unreported vertebral body misalignment errors were discovered during case reviews. False positive cases often displayed substantial image artifacts, patient rotation, and soft tissue anatomy changes. CONCLUSIONS: Our results validated the clinical utility of the EDA for bulk image reviews and highlighted the reliability and safety of IGRT, with an absolute gross patient misalignment error rate of 0.04% ± 0.02% per delivered fraction.

Localized high-risk prostate cancer harbors an androgen receptor low subpopulation susceptible to HER2 inhibition.

Patients diagnosed with localized high-risk prostate cancer have higher rates of recurrence, and the introduction of neoadjuvant intensive hormonal therapies seeks to treat occult micrometastatic disease by their addition to definitive treatment. Sufficient profiling of baseline disease has remained a challenge in enabling the in-depth assessment of phenotypes associated with exceptional vs. poor pathologic responses after treatment. In this study, we report comprehensive and integrative gene expression profiling of 37 locally advanced prostate tumors prior to six months of androgen deprivation therapy (ADT) plus the androgen receptor (AR) inhibitor enzalutamide prior to radical prostatectomy. A robust transcriptional program associated with HER2 activity was positively associated with poor outcome and opposed AR activity, even after adjusting for common genomic alterations in prostate cancer including PTEN loss and expression of the TMPRSS2:ERG fusion. Patients experiencing exceptional pathologic responses demonstrated lower levels of HER2 and phospho-HER2 by immunohistochemistry of biopsy tissues. The inverse correlation of AR and HER2 activity was found to be a universal feature of all aggressive prostate tumors, validated by transcriptional profiling an external cohort of 121 patients and immunostaining of tumors from 84 additional patients. Importantly, the AR activity-low, HER2 activity-high cells that resist ADT are a pre-existing subset of cells that can be targeted by HER2 inhibition alone or in combination with enzalutamide. In summary, we show that prostate tumors adopt an AR activity-low prior to antiandrogen exposure that can be exploited by treatment with HER2 inhibitors.

Promises, Pitfalls, and Clinical Applications of Artificial Intelligence in Pediatrics.

Artificial intelligence (AI) broadly describes a branch of computer science focused on developing machines capable of performing tasks typically associated with human intelligence. Those who connect AI with the world of science fiction may meet its growing rise with hesitancy or outright skepticism. However, AI is becoming increasingly pervasive in our society, from algorithms helping to sift through airline fares to substituting words in emails and SMS text messages based on user choices. Data collection is ongoing and is being leveraged by software platforms to analyze patterns and make predictions across multiple industries. Health care is gradually becoming part of this technological transformation, as advancements in computational power and storage converge with the rapid expansion of digitized medical information. Given the growing and inevitable integration of AI into health care systems, it is our viewpoint that pediatricians urgently require training and orientation to the uses, promises, and pitfalls of AI in medicine. AI is unlikely to solve the full array of complex challenges confronting pediatricians today; however, if used responsibly, it holds great potential to improve many aspects of care for providers, children, and families. Our aim in this viewpoint is to provide clinicians with a targeted introduction to the field of AI in pediatrics, including key promises, pitfalls, and clinical applications, so they can play a more active role in shaping the future impact of AI in medicine.

Automated detection of vertebral body misalignments in orthogonal kV and MV guided radiotherapy: application to a comprehensive retrospective dataset.

Objective. In image-guided radiotherapy (IGRT), off-by-one vertebral body misalignments are rare but potentially catastrophic. In this study, a novel detection method for such misalignments in IGRT was investigated using densely-connected convolutional networks (DenseNets) for applications towards real-time error prevention and retrospective error auditing.Approach. A total of 4213 images acquired from 527 radiotherapy patients aligned with planar kV or MV radiographs were used to develop and test error-detection software modules. Digitally reconstructed radiographs (DRRs) and setup images were retrieved and co-registered according to the clinically applied alignment contained in the DICOM REG files. A semi-automated algorithm was developed to simulate patient positioning errors on the anterior-posterior (AP) and lateral (LAT) images shifted by one vertebral body. A DenseNet architecture was designed to classify either AP images individually or AP and LAT image pairs. Receiver-operator characteristic curves (ROC) and areas under the curves (AUC) were computed to evaluate the classifiers on test subsets. Subsequently, the algorithm was applied to the entire dataset in order to retrospectively determine the absolute off-by-one vertebral body error rate for planar radiograph guided RT at our institution from 2011-2021.Main results. The AUCs for the kV models were 0.98 for unpaired AP and 0.99 for paired AP-LAT. The AUC for the MV AP model was 0.92. For a specificity of 95%, the paired kV model achieved a sensitivity of 99%. Application of the model to the entire dataset yielded a per-fraction off-by-one vertebral body error rate of 0.044% [0.0022%, 0.21%] for paired kV IGRT including one previously unreported error.Significance. Our error detection algorithm was successful in classifying vertebral body positioning errors with sufficient accuracy for retrospective quality control and real-time error prevention. The reported positioning error rate for planar radiograph IGRT is unique in being determined independently of an error reporting system.

Accelerated Hypofractionated Chemoradiation Followed by Stereotactic Ablative Radiotherapy Boost for Locally Advanced, Unresectable Non-Small Cell Lung Cancer: A Nonrandomized Controlled Trial.

Importance: Intrathoracic progression remains the predominant pattern of failure in patients treated with concurrent chemoradiation followed by a consolidation immune checkpoint inhibitor for locally advanced, unresectable non-small cell lung cancer (NSCLC). Objective: To determine the maximum tolerated dose (MTD) and use of hypofractionated concurrent chemoradiation with an adaptive stereotactic ablative radiotherapy (SABR) boost. Design, Setting, and Participants: This was an early-phase, single-institution, radiation dose-escalation nonrandomized controlled trial with concurrent chemotherapy among patients with clinical stage II (inoperable/patient refusal of surgery) or III NSCLC (American Joint Committee on Cancer Staging Manual, seventh edition). Patients were enrolled and treated from May 2011 to May 2018, with a median patient follow-up of 18.2 months. Patients advanced to a higher SABR boost dose if dose-limiting toxic effects (any grade 3 or higher pulmonary, gastrointestinal, or cardiac toxic effects, or any nonhematologic grade 4 or higher toxic effects) occurred in fewer than 33% of the boost cohort within 90 days of follow-up. The current analyses were conducted from January to September 2023. Intervention: All patients first received 4 Gy × 10 fractions followed by an adaptive SABR boost to residual metabolically active disease, consisting of an additional 25 Gy (low, 5 Gy × 5 fractions), 30 Gy (intermediate, 6 Gy × 5 fractions), or 35 Gy (high, 7 Gy × 5 fractions) with concurrent weekly carboplatin/paclitaxel. Main Outcome and Measure: The primary outcome was to determine the MTD. Results: Data from 28 patients (median [range] age, 70 [51-88] years; 16 [57%] male; 24 [86%] with stage III disease) enrolled across the low- (n = 10), intermediate- (n = 9), and high- (n = 9) dose cohorts were evaluated. The protocol-specified MTD was not exceeded. The incidences of nonhematologic acute and late (>90 days) grade 3 or higher toxic effects were 11% and 7%, respectively. No grade 3 toxic effects were observed in the intermediate-dose boost cohort. Two deaths occurred in the high-dose cohort. Two-year local control was 74.1%, 85.7%, and 100.0% for the low-, intermediate-, and high-dose cohorts, respectively. Two-year overall survival was 30.0%, 76.2%, and 55.6% for the low-, intermediate-, and high-dose cohorts, respectively. Conclusions and Relevance: This early-phase, dose-escalation nonrandomized controlled trial showed that concurrent chemoradiation with an adaptive SABR boost to 70 Gy in 15 fractions with concurrent chemotherapy is a safe and effective regimen for patients with locally advanced, unresectable NSCLC. Trial Registration: ClinicalTrials.gov Identifier: NCT01345851.

Long-Term Outcome and Complications after Transcondylar Screw Placement for Canine Humeral Intracondylar Fissure.

OBJECTIVE: The aim of this study was to report postoperative complications and long-term outcomes following transcondylar screw placement for humeral intracondylar fissure (HIF). STUDY DESIGN: It was a retrospective single-centre case series. Medical records (2018-2022) were reviewed for dogs with HIF treated with transcondylar screw placement. Data collected included signalment, concurrent orthopaedic disease, partial or complete HIF, surgeon, surgical approach, surgical technique and implant type, transcondylar screw angulation, postoperative complications and outcomes. Long-term outcome was assessed with owner questionnaire, orthopaedic examination and follow-up radiography. Statistical analysis was performed to identify risk factors predisposing to a complication or a poor outcome. RESULTS: Forty-seven dogs (57 elbows) met the inclusion criteria; long-term follow-up was available in 41 dogs (50 elbows). Minor and major medical complications were noted in seven and three elbows respectively. The total complication rate was 17.5%. Increasing age was significantly associated with a reduced risk of postoperative complications (p = 0.0051). No other risk factors were identified. A postoperative complication was not associated with a less than full outcome (p = 0.5698). CONCLUSION: Transcondylar screw placement for HIF is associated with a low complication rate and good outcome.

Identifying bias in models that detect vocal fold paralysis from audio recordings using explainable machine learning and clinician ratings.

Introduction: Detecting voice disorders from voice recordings could allow for frequent, remote, and low-cost screening before costly clinical visits and a more invasive laryngoscopy examination. Our goals were to detect unilateral vocal fold paralysis (UVFP) from voice recordings using machine learning, to identify which acoustic variables were important for prediction to increase trust, and to determine model performance relative to clinician performance. Methods: Patients with confirmed UVFP through endoscopic examination (N=77) and controls with normal voices matched for age and sex (N=77) were included. Voice samples were elicited by reading the Rainbow Passage and sustaining phonation of the vowel "a". Four machine learning models of differing complexity were used. SHapley Additive explanations (SHAP) was used to identify important features. Results: The highest median bootstrapped ROC AUC score was 0.87 and beat clinician's performance (range: 0.74 - 0.81) based on the recordings. Recording durations were different between UVFP recordings and controls due to how that data was originally processed when storing, which we can show can classify both groups. And counterintuitively, many UVFP recordings had higher intensity than controls, when UVFP patients tend to have weaker voices, revealing a dataset-specific bias which we mitigate in an additional analysis. Conclusion: We demonstrate that recording biases in audio duration and intensity created dataset-specific differences between patients and controls, which models used to improve classification. Furthermore, clinician's ratings provide further evidence that patients were over-projecting their voices and being recorded at a higher amplitude signal than controls. Interestingly, after matching audio duration and removing variables associated with intensity in order to mitigate the biases, the models were able to achieve a similar high performance. We provide a set of recommendations to avoid bias when building and evaluating machine learning models for screening in laryngology.

Increase in robot assisted operating room efficiency: A quality improvement study.

INTRODUCTION: Pyeloplasties are time-sensitive, and the most common robot assisted intervention performed in pediatric urology. Early intervention is intended to avoid permanent loss of renal function with negative long-term effects if surgery is delayed when indicated. A need to increase capacity has become a premium value in patient care. OBJECTIVE: Our aim was to reduce operative time, providing value by reducing total robotic console time in robot assisted pyeloplasty (RP) cases. We hypothesized that process improvement and supply management during RP leads to a significant reduction in operative time. METHODS: Intraoperative surgical workflow was reviewed and routine tasks performed during the various sections were selected with the goal of reducing Operating room inactivity. We focused on robotic arm activity, and total operative time to assess our outcomes. Our intervention was to standardize an OR staff task list, a priori supply inventory procurement for each anticipated major step in the case, confirmed prior to each major step. Baseline RP duration data for a single Pediatric Urologist were identified and recorded before any interventions. A clinical standard work (CSW) was developed based on optimization of equipment/supplies for the RP procedure, compartmentalized into the 8 key steps for RP. These major steps included: patient positioning, docking, retroperitoneal and ureteral dissection, hitch stitch, pyelotomy, stent placement, and anastomosis. Balancing measures included percentage trainee console use, preparatory time, and OR block start/end time. Baseline data for RP cases performed between 11/2020 and 2/2022 were automatically extracted from charts and analyzed using AdaptX (Seattle, WA). Post-intervention was between 3/2022 to 3/2023. Mann-WhitneyU was used for continuous variables for non-parametric distribution. RESULTS: 37 patients underwent RP during the study period. 15 cases were performed prior to intervention and 22 post intervention Total console time prior to intervention was 152 vs 109 min after intervention (p = 0.0002). Dual instrument inactivity was reduced from 13.1 % to 7.1 % (p < 0.0001). Dual consoles were used in 40 % vs ∼69 % pre-vs post-intervention, respectively (p = 0.5000). No difference in patient age distribution between groups was seen (p = 0.1498). Trainee operative time did not differ statistically pre- and post-intervention (63.0 vs 48.6 %, p = 0.0871). CONCLUSIONS: Decreasing surgical lapses and standardizing intraoperative tasks can result in more efficient case completion, potentially increasing OR capacity.

Impact of Interfractional Bladder and Trigone Displacement and Deformation on Radiation Exposure and Subsequent Acute Genitourinary Toxicity: A Post Hoc Analysis of Patients Treated with Magnetic Resonance Imaging-Guided Prostate Stereotactic Body Radiation Therapy in a Phase 3 Randomized Trial.

PURPOSE: Emerging data suggest that trigone dosimetry may be more associated with poststereotactic body radiation therapy (SBRT) urinary toxicity than whole bladder dosimetry. We quantify the dosimetric effect of interfractional displacement and deformation of the whole bladder and trigone during prostate SBRT using on-board, pretreatment 0.35T magnetic resonance images (MRI). METHODS AND MATERIALS: Seventy-seven patients treated with MRI-guided prostate SBRT (40 Gy/5 fractions) on the MRI arm of a phase 3 single-center randomized trial were included. Bladder and trigone structures were contoured on images obtained from a 0.35T simulation MRI and 5 on-board pretreatment MRIs. Dice similarity coefficient (DSC) scores and changes in volume between simulation and daily treatments were calculated. Dosimetric parameters including Dmax, D0.03 cc, Dmean, V40 Gy, V39 Gy, V38 Gy, and V20 Gy for the bladder and trigone for the simulation and daily treatments were collected. Both physician-scored (Common Terminology Criteria for Adverse Events, version 4.03 scale) as well as patient-reported (International Prostate Symptom Scores and the Expanded Prostate Cancer Index Composite-26 scores) acute genitourinary (GU) toxicity outcomes were collected and analyzed. RESULTS: The average treatment bladder volume was about 30% smaller than the simulation bladder volume; however, the trigone volume remained fairly consistent despite being positively correlated with total bladder volume. Overall, the trigone accounted for <2% of the bladder volume. Median DSC for the bladder was 0.79, whereas the median DSC of the trigone was only 0.33. No statistically significant associations between our selected bladder and trigonal dosimetric parameters and grade ≥2 GU toxicity were identified, although numerically, patients with GU toxicity (grade ≥2) had higher intermediate doses to the bladder (V20 Gy and Dmean) and larger volumes exposed to higher doses in the trigone (V40 Gy, V39 Gy, and V38 Gy). CONCLUSIONS: The trigone exhibits little volume change, but considerable interfractional displacement/deformation. As a result, the relative volume of the trigone receiving high doses during prostate SBRT varies substantially between fractions, which could influence GU toxicity and may not be predicted by radiation planning dosimetry.

Reducing Pediatric Posttonsillectomy Opioid Prescribing: A Quality Improvement Initiative.

OBJECTIVE: Postoperative pain is the most common morbidity associated with tonsillectomy. Opioids are frequently used in multimodal posttonsillectomy analgesia regimens; however, concerns regarding respiratory depression, drug-drug interactions, and medication misuse necessitate responsible opioid stewardship among prescribing surgeons. It is unclear if intentionally reducing opioid prescription doses negatively affects the patient experience. METHODS: A quality improvement team reviewed all posttonsillectomy opioid prescriptions at a pediatric ambulatory surgery center between January and June 2021 (preintervention, 163 patients). Following this review, we performed an opioid education session for surgeons and studied opioid prescribing habits between July and December 2021 (Plan-Do-Study-Act [PDSA] 1, 152 patients). We then implemented a standardized prescription protocol of 7 doses of oxycodone per patient and again reviewed prescriptions between January and June 2022 (PDSA 2, 178 patients). The following measures were evaluated: initial number of opioid doses prescribed, need for refills, 7-day emergency department (ED) visits, and readmissions. RESULTS: Each intervention reduced the average number of initial oxycodone doses per patient (12.2 vs 9.2 vs 6.9 doses, P < .001). There were no changes in the rate of refill requests, 7-day ED visits, and readmissions, by descriptive or Statistical Process Control analyses. DISCUSSION: In 2 PDSA cycles, we achieved a 43% reduction in the number of doses of oxycodone prescribed following tonsillectomy. We did not observe any increased rates in balancing measures, which are surrogates for unintentional effects of PDSA changes, including refills, ED presentations, and readmission rates. IMPLICATIONS FOR PRACTICE: Directed provider education and standardized posttonsillectomy prescription protocols can safely decrease postoperative opioid prescribing. Further PDSA cycles are required to consider even fewer opioid prescription doses.

Stereotactic MR-guided on-table adaptive radiation therapy (SMART) for borderline resectable and locally advanced pancreatic cancer: A multi-center, open-label phase 2 study.

BACKGROUND AND PURPOSE: Radiation dose escalation may improve local control (LC) and overall survival (OS) in select pancreatic ductal adenocarcinoma (PDAC) patients. We prospectively evaluated the safety and efficacy of ablative stereotactic magnetic resonance (MR)-guided adaptive radiation therapy (SMART) for borderline resectable (BRPC) and locally advanced pancreas cancer (LAPC). The primary endpoint of acute grade ≥ 3 gastrointestinal (GI) toxicity definitely related to SMART was previously published with median follow-up (FU) 8.8 months from SMART. We now present more mature outcomes including OS and late toxicity. MATERIALS AND METHODS: This prospective, multi-center, single-arm open-label phase 2 trial (NCT03621644) enrolled 136 patients (LAPC 56.6 %; BRPC 43.4 %) after ≥ 3 months of any chemotherapy without distant progression and CA19-9 ≤ 500 U/mL. SMART was delivered on a 0.35 T MR-guided system prescribed to 50 Gy in 5 fractions (biologically effective dose10 [BED10] = 100 Gy). Elective coverage was optional. Surgery and chemotherapy were permitted after SMART. RESULTS: Mean age was 65.7 years (range, 36-85), induction FOLFIRINOX was common (81.7 %), most received elective coverage (57.4 %), and 34.6 % had surgery after SMART. Median FU was 22.9 months from diagnosis and 14.2 months from SMART, respectively. 2-year OS from diagnosis and SMART were 53.6 % and 40.5 %, respectively. Late grade ≥ 3 toxicity definitely, probably, or possibly attributed to SMART were observed in 0 %, 4.6 %, and 11.5 % patients, respectively. CONCLUSIONS: Long-term outcomes from the phase 2 SMART trial demonstrate encouraging OS and limited severe toxicity. Additional prospective evaluation of this novel strategy is warranted.

Quantifying Intrafraction Motion and the Impact of Gating for Magnetic Resonance Imaging-Guided Stereotactic Radiation therapy for Prostate Cancer: Analysis of the Magnetic Resonance Imaging Arm From the MIRAGE Phase 3 Randomized Trial.

PURPOSE: Real-time intrafraction tracking/gating is an integral component of magnetic resonance imaging-guided radiation therapy (MRgRT) and may have contributed to the acute toxicity reduction during prostate stereotactic body radiation therapy observed on the MRgRT-arm of the MIRAGE (MAGNETIC RESONANCE IMAGING-GUIDED Stereotactic Body Radiotherapy for Prostate Cancer) randomized trial (NCT04384770). Herein we characterized intrafraction prostate motion and assessed gating effectiveness. METHODS AND MATERIALS: Seventy-nine patients were treated on an MR-LINAC. Real-time cine imaging was acquired at 4Hz in a sagittal plane. If >10% of the prostate area moved outside of a 3-mm gating boundary, an automatic beam hold was initiated. An in-house tool was developed to retrospectively extract gating signal for all patients and identify the tracked prostate in each cine frame for a subgroup of 40 patients. The fraction of time the prostate was within the gating window was defined as the gating duty cycle (GDC). RESULTS: A total of 391 treatments from 79 patients were analyzed. Median GDC was 0.974 (IQR, 0.916-0.983). Fifty (63.2%) and 24 (30.4%) patients had at least 1 fraction with GDC ≤0.9 and GDC ≤0.8, respectively. Incidence of low GDC fractions among patients appeared stochastic. Patients with minimum GDC <0.8 trended toward more frequent grade 2 genitourinary toxicity compared with those with minimum GDC >0.8 (38% vs 18%, P = .065). Prostate intrafraction motion was mostly along the bladder-rectum axis and predominantly in the superior-anterior direction. Motion in the inferior-posterior direction was associated with significantly higher rate of acute grade 2 genitourinary toxicity (66.7% vs 13.9%, P = .001). Gating limited mean prostate motion during treatment delivery in fractions with a GDC <0.9 (<0.8) to 2.9 mm (2.9 mm) versus 4.1 mm (4.7 mm) for ungated motion. CONCLUSIONS: Fractions with large intrafraction motion were associated with increased toxicity and their occurrence among patients appears stochastic. Real-time tracking/gating effectively mitigated this motion and is likely a major contributing factor of acute toxicity reduction associated with MRgRT.